Left Ventricular Noncompaction (Left Ventricular Hypertrabeculation)

A morphological trait of the left ventricular myocardium — prominent trabeculations and deep intertrabecular recesses — occurring as a physiological variant, an adaptive response to loading, or with sarcomeric, neuromuscular, mitochondrial and syndromic cardiomyopathies; reclassified by the European Society of Cardiology in 2023 as a phenotypic trait rather than an independent cardiomyopathy.

Literature and regulatory information updated: 6 October 2026.

0. Initial Reference Table

1. Introduction

1.1. Historical overview

Clinical Cardiogenetics attributes the first description of an abnormal ventricular myocardial pattern, with prominent trabeculations and deep recesses, to Feldt and colleagues, in 1969, in a paediatric case of complex congenital heart disease [1, 25]. The French PNDS goes back to Grant, who in 1926 described a “spongy myocardium” in a heart with other congenital malformations, and places in 1984 the first echocardiographic description of an isolated form, then called persistence of myocardial sinusoids [23]. In infants with structural heart disease, the deep endomyocardial spaces, then called sinusoids, were attributed to excessive intracavitary pressures during embryonic and fetal development [4]. In 1990, Chin and colleagues brought together eight patients with “isolated noncompaction of left ventricular myocardium”, i.e. without other structural heart disease, and proposed the first quantitative echocardiographic criterion: the ratio of the distance from the epicardium to the trough of the trabecular recess to the distance from the epicardium to the peak of the trabeculation (X/Y) of 0.5 or less [1, 26]. The designation “noncompaction” carried an explicit pathogenic hypothesis — arrest of compaction of a spongy embryonic myocardium, which at weeks 8 to 10 of embryonic life is still predominantly trabeculated [1].

In 2001, Jenni and colleagues in Zurich compared hearts from seven patients with isolated noncompaction with ten explanted hearts with dilated cardiomyopathy (DCM) and nine with left ventricular hypertrophy, and defined four echocardiographic criteria, among which the noncompacted/compacted (NC/C) ratio threshold above 2 at end-systole derived from the maximum value observed in the DCM hearts [1, 27]. The article’s subtitle — “a step towards classification as a distinct cardiomyopathy” — anticipated the decision of the American Heart Association (AHA), which included left ventricular noncompaction among primary cardiomyopathies of genetic cause in 2006 [1, 8]. In its 2008 classification, the European Society of Cardiology (ESC) retained it among unclassified cardiomyopathies because of a lack of consensus on whether it was an independent disease or a trait shared by different cardiomyopathies and congenital heart diseases [1, 14].

1.2. From “cardiomyopathy” to “phenotypic trait”

The widespread use of high-resolution echocardiography and cardiac magnetic resonance (CMR) showed that noncompaction criteria are fulfilled by a substantial proportion of people without cardiac disease. In MESA (Multi-Ethnic Study of Atherosclerosis), 43% of asymptomatic participants without cardiac disease or hypertension fulfilled the Petersen criterion in at least one segment; in an adult meta-analysis, pooled prevalence was 14.79% in cohorts studied by CMR and 1.28% in those studied by echocardiography [4, 13]. Pregnancy, endurance training, chronic haemolytic anaemias and chronic renal failure can induce reversible trabeculation [4, 28]. In 2017, Hershberger and colleagues concluded that the evidence pointed to a phenotype rather than a disease, a position reiterated in the GeneReviews note on DCM, according to which the relationship of the finding to the presence or severity of DCM remains unknown [21, 29]. In 2018, Ross and Semsarian warned against overdiagnosis of an entity that “we do not understand” [21].

Developmental biology removed the foundation of the designation itself. The JACC: Cardiovascular Imaging expert panel (2023) reviewed the morphometric and proliferation-labelling evidence: the trabeculated and compact layers both grow throughout embryonic and fetal life, the compact layer grows faster (allometric growth), and there is no evidence that trabeculae “coalesce” to form the compact wall; it recommended abandoning the term “noncompaction” in favour of “excessive trabeculation” [4]. The 2023 ESC Guidelines for cardiomyopathies followed the same line: the task force stated that it does not consider left ventricular noncompaction a cardiomyopathy in the general sense, but a phenotypic trait that may occur in isolation or in association with other developmental abnormalities, hypertrophy, dilatation or systolic dysfunction, and recommended the term “hypertrabeculation”, particularly when the phenomenon is transient or of adult onset [3]. The same guidelines created the category of non-dilated left ventricular cardiomyopathy (NDLVC), defined by non-ischaemic scar or fatty replacement of the left ventricle, with or without wall motion abnormalities, or by isolated global hypokinesia without scar — a category into which some patients previously labelled as having “noncompaction cardiomyopathy” with dysfunction or scar without dilatation fall [3].

This reclassification is not universal. The 2019 Heart Rhythm Society (HRS) consensus on arrhythmogenic cardiomyopathy still describes left ventricular noncompaction as an autosomal dominant genetic disease, with nine phenotypic forms and its own recommendations for diagnosis, implantable cardioverter-defibrillator (ICD) and anticoagulation [20]. The 2019 AHA scientific statement on cardiomyopathy in children retains a section dedicated to “LVNC cardiomyopathy” and stresses that, in the paediatric age group, severe phenotypes and genetic associations make the approach different from that in adults [14]. The French PNDS (HAS, 2026) records the position of ESC 2023 and the preference for the term “hypertrabeculation”, but continues to treat left ventricular noncompaction (NCVG) as a cardiomyopathy with its own recommendations for diagnosis, family screening and follow-up, and concludes, from the genetic spectrum of the French cohort (genes shared with DCM, with overrepresentation of HCN4), that it is an autonomous genetic disease [23]. In 2026, an editorial in the Journal of the American College of Cardiology argued, in its very title, for retiring the diagnosis of LVNC as conceptually flawed [30]. The Clinical Cardiogenetics chapter (2020), written in the midst of the transition, describes left ventricular noncompaction as a morphological trait and reserves the designation “LVNC cardiomyopathy” for when the trait is accompanied by DCM, a situation it considers difficult to diagnose and of genetic cause in most cases [1].

1.3. Scope

This article treats left ventricular hypertrabeculation as a phenotype that requires three successive questions: is the finding physiological or adaptive? Is there an associated cardiomyopathy (dilated, hypertrophic, restrictive, arrhythmogenic or NDLVC)? Is there a genetic cause, isolated or syndromic? The associated cardiomyopathies (hypertrophic cardiomyopathy and familial dilated cardiomyopathy) and Barth syndrome enter the differential diagnosis; only the aspects that change when excessive trabeculation is present are presented here. The historical term VENC/LVNC is retained when it reproduces the terminology of the sources, particularly series and consensus statements prior to 2023.

2. Molecular Genetics and Pathophysiology

2.1. Genetic architecture: a shared phenotype

Left ventricular hypertrabeculation is genetically heterogeneous, and most implicated genes are shared with hypertrophic cardiomyopathy (HCM), DCM and restrictive cardiomyopathy (RCM), sometimes with identical variants in patients with different phenotypes [1, 31]. The Clinical Cardiogenetics chapter listed 27 genes associated with the isolated form and estimated that sarcomeric variants explained approximately 30% of patients, with MYH7 accounting for up to 21% of isolated forms [1]. In the first large sarcomeric study, Klaassen and colleagues found heterozygous variants in 11 of 63 adult probands with isolated LVNC — seven in MYH7, one in ACTC1 and one in TNNT2, among nine distinct variants — with segregation in four autosomal dominant families and predominant localisation at the ATP-binding site of the β-myosin heavy chain [32].

The most recent series, with extended panels, have changed the hierarchy. In the Dutch registry of 327 unrelated probands (4 centres), 32% had an identified variant, 16% had familial cardiomyopathy without a variant and 52% were sporadic; MYH7, MYBPC3 and TTN accounted for 71% of variants [9]. In a German registry of 95 patients with exome sequencing, classification according to the American College of Medical Genetics and Genomics (ACMG) rules identified TTN, LMNA and MYBPC3 as the most prevalent genes; 13 patients had a pathogenic truncating TTN variant (odds ratio 40.7; confidence interval [CI] 21.6–76.6), located in exons with a transcript inclusion index (percent spliced in) of 76–100%, and a titin-related mechanism was found in one in four patients [33]. In the Kayvanpour meta-analysis (nine genetic studies in adults), mean frequencies were 11% for TTN, 9% for MYH7, 5% for MYBPC3, 3% for CASQ2 and LDB3, 2% for TPM1, MIB1, ACTC1 and LMNA and 1% for PLN, HCN4, TAZ, DTNA, TNNT2 and RBM20 [10]. In the nine genetic studies of that meta-analysis, with different panels and interpretation criteria, the proportion of patients with a pathogenic or likely pathogenic variant ranged from 38% to 59% [23].

In the French prospective multicentre study, subsequent to the meta-analysis, 95 unrelated adults were studied with a panel of 107 cardiomyopathy and arrhythmia genes: 52 pathogenic or likely pathogenic variants were identified in 42% of patients, nine of whom (9.5%) had more than one variant; the most frequent genes were TTN, followed by HCN4, MYH7 and RYR2, and the distribution included 13 genes already described in LVNC and 10 candidate genes [23]. By functional class, sarcomeric genes (56%) and ion channel genes (21%) predominated, followed, less frequently, by cytoskeletal, desmosomal and transcription factor genes [23]. The PNDS estimates a genetic cause in at least 30–50% of patients, with variants described in more than 110 nuclear and mitochondrial genes, and highlights, in the forms associated with congenital heart disease, disruption of the NOTCH pathway as a common pathway, alongside mitochondrial and metabolic defects [23].

The rare variant burden analysis by Mazzarotto and colleagues (840 cases from six cohorts — Italy, the United States, the Netherlands, France, Poland/USA and Switzerland/Germany — compared with 125,748 gnomAD controls) concluded that LVNC is, in most cases, a phenotypic variation of DCM or HCM, but identified variant classes specifically associated with LVNC: truncating variants in MYH7, ACTN2 and PRDM16 [34]. Truncating MYH7 variants, usually considered non-pathogenic for cardiomyopathy, were enriched 20-fold in cases, and carriers identified in UK Biobank and healthy volunteers had more trabeculation than matched controls; RYR2 exon deletions and transmembrane HCN4 variants were also enriched, consistent with the reported arrhythmic phenotypes [34]. In UK Biobank, trabecular complexity measured by fractal dimension has a heritability of approximately 20%, and the genome-wide association study identified 16 loci involved in cytoskeletal arborisation, some containing Mendelian cardiomyopathy genes (TTN, TNNT2, PLN); alleles associated with lower trabecular complexity conferred a greater risk of DCM and heart failure [4].

2.2. Associated genes, individually

Table 2.1 brings together the genes with reported associations, distinguishing the strength of the evidence. The percentages are not interchangeable: they depend on the population (children or adults), the imaging criterion and the panel used.

Table 2.1 — Genes associated with hypertrabeculation/LVNC

The genetic appendix of the French PNDS attributes about 10–19% of cases to TTN (mainly truncating variants), 3–13% to MYH7, 4–8% to MYBPC3 and 1–4% to RBM20, and adds to the genes of Table 2.1 the transcription factors TBX20 (variants associated with LVNC and DCM) and GATA4 (associated with septal defects and patent ductus arteriosus) [23]. The same appendix groups DTNA with the nuclear envelope genes; α-dystrobrevin, however, belongs to the dystrophin–glycoprotein complex (Table 2.1) [7, 23].

The 2022 consensus table contains two locus errors (LDB3 is given as 11p15.1 and TAZ as Xp28); the loci in the table follow the OMIM phenotypic series and, for TAFAZZIN, GeneReviews. The text of OMIM entry 601493 states “10p23”, but the gene–phenotype table in the same entry gives 10q23.2, which is the location adopted [5, 6, 11].

OMIM phenotypic series PS604169 organises LVNC into ten numbered phenotypes (Table 2.2). Nine have a known molecular basis (mapping key 3) and autosomal dominant inheritance; in seven, the same phenotype number also designates a form of DCM, and 601493 additionally includes HCM 24, translating the phenotypic overlap described above into nomenclature [6]. LVNC2 (609470) is only a locus at 11p15, mapped by linkage in a family with isolated autosomal dominant LVNC (maximum LOD 5.06; critical interval 6.8 Mb), without an identified gene; the positional candidates CSRP3 and SOX6 had no causal variants [6]. The series does not include TTN or LMNA, despite their being among the most frequent genes in clinical series, and the LVNC1 entry notes that LVNC also occurs in Barth syndrome (TAFAZZIN, Xq28) [6, 10].

Table 2.2 — OMIM phenotypic series PS604169 (accessed 7 October 2026)

Source: [6].

2.3. Complex genotypes, correlations and penetrance

A Dutch study preceding the multicentre registry found multiple variants or compound/double heterozygosity in 25% of children and 10% of adults with LVNC, a proportion higher than the 7% described in HCM; the presence of several sarcomeric variants, particularly truncating variants, was associated with a more severe phenotype and onset in childhood, and adults with several variants had more symptoms than carriers of a single variant [1]. Variants were more frequent in children and were associated with major adverse cardiac events; the risk of systolic dysfunction was higher in patients with multiple variants and with TTN variants, whereas carriers of MYH7 variants had a low risk of events [9]. In the family study, variants in the MYH7 head were associated with isolated LVNC and a more benign course, and variants in the MYH7 tail with LVNC with dilatation, systolic dysfunction and DCM without LVNC in relatives; LVNC with hypertrophy (4%) was associated with MYBPC3, and 37% of variant-carrying relatives had no phenotype [36]. In the German registry, LMNA and RBM20 variants were associated with a worse outcome [33]. In the French cohort, patients with a variant in an ion channel gene (including RYR2 and HCN4) had a higher median left ventricular ejection fraction (60% versus 40%; p <0.001) and more frequent biventricular noncompaction (53% versus 18%; p <0.001), and carriers of HCN4 variants had sinus bradycardia, sometimes marked [23]. The PNDS states that TTN variants might be associated with a lower risk of major cardiovascular events, which diverges from the Dutch registry, in which TTN was associated with more systolic dysfunction; the sources themselves acknowledge that genotype–phenotype correlations have not been consistently reproduced [9, 23]. The absence of a variant in about half of the familial forms has been attributed to errors of phenotype assignment, genes not yet identified, regions not analysed and the incomplete sensitivity of the methods [1].

2.4. Cellular and tissue mechanisms

Clinical Cardiogenetics describes four mechanistic pathways: deficient force generation due to sarcomeric variants (through a dominant negative effect, in which the mutant protein acts as a “poison polypeptide”, or through haploinsufficiency); defective force transmission due to cytoskeletal variants; energy deficit due to variants in ATP-regulating genes; and altered calcium homeostasis, through altered availability or myofibrillar sensitivity [1]. In this reading, trabeculation would be a compensatory response to the underlying dysfunction, and the location of the variant and a dose effect would determine the final phenotype [1]. Growth and differentiation pathways also participate: in mice, myocardial inactivation of Mib1 reduces NOTCH1 activation and reproduces the phenotype; PRDM16 haploinsufficiency in zebrafish causes contractile dysfunction, partial electrical uncoupling and reduced cardiomyocyte proliferation, and the nonsense allele p.Gln187Ter (published designation PRDM16-Q187X; the source gives neither the c. notation nor the transcript) causes deficient myocardial development with altered TGF-β signalling and noncompaction in humans and mice; inactivation of myocardial development genes (PBP, JMJ, FKBP12, Sp3, NKX2-5, BMP10) causes congenital noncompaction in mice, but, with the exception of NKX2-5, no variants in these genes have been found in humans [1, 38, 40, 44]. The only induced pluripotent stem cell (iPS)-derived cardiomyocyte model of LVNC published up to 2020, with a TBX20 variant, showed reduced baseline proliferative capacity [45].

The classical “arrest of compaction” hypothesis was based on the similarity between LVNC myocardium and the embryonic myocardium of weeks 8 to 10 [1]. Three observations contradict it: the trabeculated and compact layers both grow, with faster growth of the compact layer; the compact layer forms normally even when excessive trabeculation is induced by suppression of NKX2-5; and the trabecular cardiomyocytes of patients with LVNC have already lost their embryonic molecular identity [3, 4]. The ESC Guidelines add that acquired and transient hypertrabeculation (athletes, pregnancy, intense physical activity) must correspond to greater prominence of a normal architecture, because cardiomyocytes are terminally differentiated and cannot form new structures [3]. The French PNDS retains the embryological hypothesis as the explanation for the designation, placing compaction between the 5th and 8th week, and not between the 8th and 10th as Clinical Cardiogenetics does; it adds that, in the embryo, the compact layer mainly ensures contraction and the trabeculae rapid and coordinated impulse conduction, giving rise to the ventricular conduction system, so a functional defect of the trabeculae could contribute to the diversity of manifestations [1, 23]. The same document acknowledges that the arrest hypothesis explains neither the partial or complete regression of trabeculation nor its occurrence in pregnancy, in elite athletes and in sickle cell disease [23].

2.5. Pathological anatomy

Macroscopically, the noncompacted endocardial layer contains numerous prominent trabeculations, with deep recesses penetrating the compact layer, particularly in the apical and mid-segments of the inferior and lateral walls [1]. In Burke’s series (14 paediatric cases), three patterns were described — broad anastomosing trabeculae; coarse trabeculae resembling multiple papillary muscles; and fine interwoven muscle bundles or a relatively smooth surface with compressed invaginations, identified mainly microscopically — without morphological differences between isolated and non-isolated forms [1]. All 14 hearts had endocardial fibrosis with elastin deposition, 10 had subendocardial replacement fibrosis consistent with ischaemic microinfarcts and 6 had right ventricular involvement. Another study showed endocardium lining the recesses in continuity with the cavity, ischaemic lesions in the thickened endocardium and trabeculae, absent to severe interstitial fibrosis, no fibre disarray and, in some patients, chronic inflammation and abnormalities of intramyocardial vessels [1]. Freedom and colleagues proposed two pathological criteria: absence of well-formed papillary muscles and histological penetration of endocardial recesses through more than 50% of the thickness towards the epicardium; the recesses do not communicate with the coronary circulation [1]. In an autopsy study of 474 normal hearts, prominent trabeculations were present in 68%, but more than three trabeculations were present in only 3.4% [1].

2.6. Barth syndrome: the metabolic model

Barth syndrome is caused by loss-of-function variants in TAFAZZIN (formerly TAZ, Xq28), which encodes an inner mitochondrial membrane transacylase responsible for remodelling the acyl chains of immature cardiolipin into tetralinoleoyl cardiolipin [5, 46]. Deficiency reduces mature cardiolipin and causes the accumulation of monolysocardiolipin, disturbing mitochondrial structure, respiratory supercomplex organisation and apoptosis [5]. Almost 45% of the genomic sequence of TAFAZZIN consists of interspersed repetitive sequences (76% of which are Alu), which may explain recurrent de novo rearrangements [5]. Cardiomyopathy — dilated, with or without endocardial fibroelastosis, with frequent hypertrabeculation and, less often, hypertrophic — almost always presents before five years [5].

3. Clinical Spectrum and Manifestations by System

3.1. Phenotypic spectrum

Presentation ranges from an asymptomatic incidental finding to severe heart failure requiring transplantation, with marked variability even within the same family [1]. Age at diagnosis ranges from the prenatal period to 94 years; in isolated adult forms, the median lies between the fourth and fifth decades [1]. In the meta-analysis of 35 studies involving 2,271 adult patients, the mean age at diagnosis was the mid-fifth decade, two-thirds were men, 7% had congenital heart disease (an atrial or ventricular septal defect or Ebstein anomaly), 24% had a family history of cardiomyopathy and 5% had neuromuscular disease [10]. The literature review cited in Clinical Cardiogenetics found a male:female ratio close to 2:1, which was not explained by X-linked forms [1].

The 2019 HRS consensus describes nine forms: the benign form (normal dimensions, thickness and systolic and diastolic function, without early arrhythmias), the right ventricular form, biventricular form, dilated form, hypertrophic form, restrictive form, mixed form (hypertrophic and dilated, or dilated and restrictive), the form with congenital heart disease and the arrhythmogenic form; the myocardium may change from one form to another (“undulating phenotype”) [20]. Arbustini proposed seven groups: idiopathic LVNC with normal function and dimensions; LVNC with dilatation and dysfunction from the outset; LVNC fulfilling criteria for arrhythmogenic cardiomyopathy, DCM, HCM or RCM; LVNC with congenital heart disease; syndromic LVNC; acquired and potentially reversible LVNC; and right ventricular noncompaction, with or without LVNC [1]. In the 113 Dutch families studied by van Waning and colleagues, 43% of probands had isolated LVNC, 53% LVNC with dilatation and 4% LVNC with hypertrophy [36].

3.2. Heart failure and ventricular dysfunction

Heart failure is the most frequent presentation in clinical series, with dyspnoea, fatigue, atypical chest pain and presyncope or syncope; LVNC may also manifest as peripartum cardiomyopathy [1]. The PNDS estimates that heart failure affects up to 70% of patients in clinical series, with dyspnoea, exertional fatigue and, in severe cases, pulmonary oedema, and states that the disease may be revealed by family screening, by palpitations or syncope, or by embolism from an intracavitary thrombus [23]. Systolic dysfunction is the main mediator of prognosis: in the cohorts brought together by the expert panel, the aggregate cardiovascular mortality of patients with reduced ejection fraction was double that of patients with normal function [4]. In children, the course may be biphasic: among 36 children at Texas Children's Hospital (median age 90 days), 83% had depressed ejection fraction at diagnosis (median 30%), nine infants transiently recovered function and deteriorated again after a median of 6.3 years, and two showed an undulating phenotype from dilated to hypertrophic [47].

3.3. Arrhythmias and conduction disturbances

No arrhythmia is characteristic or pathognomonic [1]. Table 3.1 brings together the abnormalities described in Clinical Cardiogenetics; in the adult meta-analysis, conduction disease was present in 26%, supraventricular tachycardia in 17% and sustained or non-sustained ventricular tachycardia (VT) in 18% [1, 10]. The PNDS gives a prevalence of arrhythmias that may reach 40%, mainly atrial fibrillation and VT, and bundle branch blocks as the most frequent conduction disturbances; in the first French registry of 105 adults with isolated LVNC, the most frequent electrocardiographic abnormalities were left ventricular hypertrophy (18%), complete left bundle branch block (15%) and atrial fibrillation (6%) [23].

Table 3.1 — Arrhythmias and conduction disturbances described in LVNC

Source: Table 7.2 of Clinical Cardiogenetics [1].

In Bhatia’s systematic review cited by the HRS consensus, the prevalence of VT approached 40%, and sudden death accounted for more than 55% of LVNC-related deaths [20]. In an American series of 242 children with isolated LVNC (1990–2009), 33.1% had arrhythmias, 17.4% VT and 6.2% sudden death; arrhythmia was associated with greater mortality (hazard ratio [HR] 2.8), and 14 of the 15 sudden deaths occurred in children with abnormal dimensions or function [48]. Some gene-specific associations guide investigation: complete AV block with an unfavourable prognosis, particularly in the fetus; intraventricular conduction disturbances with ZASP (LDB3) variants; juvenile atrial fibrillation with ACTN2; and sinus dysfunction with HCN4 [14]. In an observational analysis of electronic health records (TriNetX platform), the incidence of atrial fibrillation was 36 per 1,000 person-years during the first year after LVNC diagnosis, and atrial fibrillation was associated, after propensity score matching, with more stroke (HR 1.47), de novo heart failure (HR 1.44) and death (HR 1.26); identification through diagnostic codes does not establish which imaging criteria were applied [49].

3.4. Thromboembolism

Thromboembolism includes stroke or transient ischaemic attack, peripheral embolism and mesenteric thrombosis [1]. The mean frequency of systemic thromboembolic events was 9% in the adult meta-analysis [10]. The risk attributable to trabeculation, however, is debatable: the long-term follow-up studies cited in Clinical Cardiogenetics found no association between the presence or extent of hypertrabeculation and thromboembolic risk, and the expert panel considers ventricular thrombus to be associated with the myocardial dysfunction and heart failure that coexist with trabeculation [1, 4]. In an international multicentre study of 1,160 patients with DCM assessed by CMR (22 centres, mainly Spanish and Dutch; genetic testing in 86%), hypertrabeculation was present in approximately 30%. Over a median follow-up of 5.1 years, embolic events occurred in 3.2%, without an association with hypertrabeculation (HR 1.5; 95% CI 0.75–3.00), even in patients in sinus rhythm with an ejection fraction ≤40% (HR 1.89; 95% CI 0.7–5.5), whereas atrial fibrillation and reduced ejection fraction were associated with embolism [16].

3.5. Non-isolated forms and extracardiac manifestations

Congenital heart disease. The association predominates in children: in one study, 78% of 46 children with LVNC had congenital heart disease [1]. The most frequent lesions in the series collated in Clinical Cardiogenetics were atrial septal defect (22/135; 16%), ventricular septal defect (23/218; 11%), patent ductus arteriosus (16/182; 9%) and Ebstein anomaly (11/130; 8%, in addition to 14 isolated cases), followed by left atrial appendage isomerism, bicuspid aortic valve, pulmonary stenosis and atresia, aortic coarctation, hypoplastic left heart syndrome and valvular dysplasias [1]. Sarcomeric variants have been described in families with cardiomyopathy and malformations (septal defects, Ebstein anomaly, patent ductus arteriosus, tetralogy of Fallot and coarctation), but more than one member of a family rarely has the malformation [1].

Neuromuscular disease. Stöllberger and Finsterer described excessive trabeculation in Duchenne and Becker muscular dystrophies and myotonic dystrophy; in a prospective cohort, almost 80% of patients with hypertrabeculation examined by a neurologist had neuromuscular disease, which was an independent predictor of death [1, 4]. In a large cohort of dystrophinopathies, 20% had cardiomyopathy with excessive trabeculation according to echocardiographic criteria; on CMR, almost 30% of patients with Duchenne muscular dystrophy had excessive trabeculation in at least one segment, with an increase in the trabeculated/compact ratio of +0.4 per year [4]. GeneReviews on Barth syndrome reports a prevalence of LVNC of up to 28% in Duchenne muscular dystrophy, where, unlike Barth syndrome, the finding tends to worsen [5]. Other neuromuscular associations include laminopathies, “zaspopathy” (LDB3), dystrobrevinopathy and myoadenylate deaminase deficiency [1, 4].

Mitochondrial disease. In a series of 113 children with mitochondrial disease, 13% had LVNC; among 36 children with LVNC who underwent muscle biopsy, five had morphological and biochemical evidence of a mitochondrial defect, including partial deficiency of respiratory chain complexes I–III [1].

Barth syndrome. Cardiomyopathy is the most severe manifestation, present in 73% of presentations during the first year of life (cardiac findings), generally dilated, with frequent hypertrabeculation; it is accompanied by neutropenia (18% of findings during the first year), proximal skeletal myopathy, prepubertal growth delay with postpubertal catch-up, and 3-methylglutaconic aciduria [5, 46]. Dysfunction may improve by mid-childhood, but 25% of children may require transplantation in early childhood; left or right ventricular noncompaction, prolonged QTc, right ventricular dysfunction and the restrictive phenotype are risk factors for cardiac death or transplantation [5]. Approximately 13% of adults require an ICD because of ventricular arrhythmia [46].

3.6. Fetal and neonatal forms

Prenatal diagnosis more frequently detects biventricular hypertrophy or hypertrabeculation than the typically left-sided abnormalities of adults [1]. Fetal forms may cause hydrops of cardiac origin, as in the case described in section 10, and have an unfavourable prognosis when associated with complete AV block [14, 50]. The expert panel recommends caution in children with excessive trabeculation, even with normal function, because of the reported association with genetic and neuromuscular diseases [4].

3.7. Particular features of the paediatric age group

The French PNDS distinguishes two clinical forms in children: the form with left ventricular dilatation and systolic dysfunction and the form without dilatation or systolic dysfunction, but with diastolic dysfunction and left atrial dilatation, close to the restrictive forms [23]. The restrictive forms are the most severe, followed by the dilated forms; the isolated forms have the best prognosis, followed by those associated with hypertrophy [23]. In infants, particularly when the myocardium is hypertrophic, a metabolic cause should be suspected first (mitochondrial disease, Barth syndrome in boys or another inborn error of metabolism), found in 25% of infant forms; very high QRS voltages and a short PR interval on the ECG support that hypothesis [23]. Noncompaction is frequent in congenital heart disease, particularly in Ebstein anomaly, but also in tetralogy of Fallot and aorto-left ventricular tunnel; in these associated forms, the yield of genetic panels is higher than in the isolated forms [23]. Paediatric forms with a dilated or hypertrophic phenotype or with congenital heart disease result more often from multiple variants or de novo variants [23].

4. Clinical Recognition and Warning Signs

In most adults, recognition begins with an echocardiography or CMR report describing excessive trabeculation. The first clinical task is to distinguish the finding from myocardial disease. The 2023 expert panel considers that, in adults with incidental trabeculation and normal function and morphology, management is determined by other cardiovascular symptoms or abnormalities, without regard to the trabecular pattern, and that these patients appear to require little or no long-term follow-up unless a family history or electrocardiographic abnormalities are present [4]. The HRS consensus emphasises that a normal electrocardiogram (ECG) is rare in LVNC with disease, being abnormal in 80–90% of cases; increased voltage in the anterolateral leads is common in young children, sometimes with pre-excitation [20].

Table 4.1 — Signs shifting excessive trabeculation from “probable variant” to “probable disease”

In athletes, the expert panel recommends concern only when ventricular dysfunction or dilatation, symptoms, an abnormal ECG unexplained by training or a family history of cardiomyopathy is present; in these cases, it recommends cardiopulmonary exercise testing, Holter monitoring and CMR [4]. Since 2013, academic sports cardiology centres in England have stopped investigating asymptomatic athletes with normal function and ECG findings who fulfil echocardiographic criteria for excessive trabeculation [4].

In the paediatric setting, the 2023 ESC guidelines propose an approach based on age at onset, consanguinity and family history, cardiac and systemic involvement, ECG, imaging and laboratory assessment; in infants, the investigation should include hypotonia and weakness, hypoglycaemia, acidosis, CK and transaminases, ketonuria, organic acids, acylcarnitine profile and calcium and vitamin D metabolism, with a multidisciplinary team of genetics, metabolic diseases and neurology [3]. The French PNDS specifies the fasting metabolic sampling in infants: blood lactate, CK, plasma acylcarnitine profile, urinary organic acid chromatography, ammonia and blood glucose, with a full blood count to look for neutropenia [23].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Imaging criteria

There is no reference standard; the criteria were derived from small retrospective series without external validation, and diagnosis by imaging alone has limited specificity [1, 4]. Table 5.1 summarises the most frequently cited criteria.

Table 5.1 — Imaging criteria for excessive trabeculation/LVNC

Sources: Tables 7.1 and 7.3 of Clinical Cardiogenetics, Table 1 of the 2023 expert panel and Table 5 of the HRS consensus [1, 4, 20]. For Captur’s fractal criterion, the expert panel’s table gives a threshold of “≥1.30”, different from the value of 1.392 presented by Clinical Cardiogenetics; the discrepancy probably relates to the ventricular region considered and was not resolved in the sources consulted [1, 4].

The Jenni threshold was set at 2.0 because this was the maximum ratio found in the hearts with DCM in the derivation study [1]. When applied to large populations, the criteria lost specificity: up to 23% of patients with DCM fulfilled echocardiographic criteria, and 4–13% of asymptomatic low-risk populations fulfilled the CMR criteria of Petersen, Jacquier and Stacey [1]. In a study of 199 patients with systolic dysfunction, 23.6% fulfilled echocardiographic criteria, as did five healthy controls, four of whom were black [20]. In an analysis of 700 patients referred for CMR, prevalence varied widely according to the criterion, and none was associated with death, ischaemic stroke, VT, ventricular fibrillation or hospitalisation for heart failure [20]. The HRS consensus therefore assigns only class IIb (level B-NR) to the use of the NC/C ratio to establish the diagnosis and to CMR or other advanced imaging for diagnosis and risk stratification when ventricular arrhythmias are present; on computed tomography, a ratio >2.3 distinguished LVNC with a sensitivity of 88% and specificity of 97% in a series of eight patients [20]. The ESC guidelines state that echocardiographic contrast agents may be considered to improve delineation of the endocardium, demonstrate hypertrabeculation or exclude intraventricular thrombus [3].

In practice, a diagnosis of disease rests on integration with systolic and diastolic function, dimensions, tissue characterisation by CMR (LGE), ECG, rhythm, family history and loading context, rather than the NC/C ratio alone [3, 4]. In the ESC guidelines, hypertrabeculation is not among the morphological traits defining cardiomyopathy phenotypes (hypertrophy, dilatation, non-ischaemic scar, systolic and diastolic dysfunction); the final diagnosis should describe the aetiology alongside the phenotype [3].

The French PNDS retains echocardiography as the first-line investigation, with three criteria (multiple trabeculations with an NC/C ratio >2, intertrabecular recesses perfused on colour Doppler and absence of associated heart disease), and considers CMR the second baseline investigation, to be performed in all patients with suspected noncompaction, because it confirms the diagnosis, assesses extent, detects fibrosis and thrombi and has prognostic value [23]. It stresses the poor agreement between the three echocardiographic definitions and the two CMR definitions in the same patient and recommends, in doubtful forms, multimodality imaging and the opinion of a specialised centre [23]. The systematic indication for CMR diverges from the class IIb assigned by the 2019 HRS consensus to advanced imaging and from the position of the 2023 panel, which does not provide for further investigation in the adult with an incidental finding and normal function and morphology [4, 20].

5.2. Initial assessment of the proband

Clinical Cardiogenetics proposes an assessment of the proband adapted from the ACCF/AHA heart failure guidelines (Table 5.2).

Table 5.2 — Proposed assessment of the proband with LVNC

Source: Table 7.9 of Clinical Cardiogenetics [1]. The PNDS adds to the initial assessment Holter, exercise testing, exercise echocardiography or cardiopulmonary exercise testing with oxygen consumption (VO₂) to measure functional capacity, and periodic measurement of B-type natriuretic peptide (BNP) or its N-terminal fragment (NT-proBNP) as a marker of heart failure severity [23].

5.3. Genetic testing: indications

The recommendations differ regarding the population to test and converge on the principle of testing the disease phenotype rather than the imaging finding (Table 5.3).

Table 5.3 — Recommendations on genetic testing in LVNC/hypertrabeculation

Sources: [3, 11, 20, 23].

The 2022 consensus justifies the restriction: incidental LVNC in pregnancy, in athletes or on imaging has a high prevalence in the adult population, and testing in those settings may cause more harm than benefit, mainly because of variants of uncertain significance (VUS) [11]. The DCM GeneReview, for its part, recommends offering testing to all people with non-ischaemic DCM of any age, including peripartum, with a multigene panel restricted to genes with definitive, strong or moderate evidence in ClinGen — a recommendation that applies to the patient with hypertrabeculation and a dilated phenotype [21]. The PNDS does not exclude the isolated finding with normal function from the indication for testing in the index case, unlike the 2022 consensus, and refers practical aspects (who may prescribe, interpretation, variant classification and the legal framework in France) to the 2024 expert consensus of the CARDIOGEN filière on counselling and genetic testing in inherited heart diseases [11, 23].

5.4. Laboratory strategy, yield and limitations

The yield ranges from 17% to 41% in LVNC series, is very low in isolated LVNC without dysfunction detected incidentally and is higher when there is another cardiomyopathy, ventricular dysfunction, syndromic features or a strong family history [11]. In adult series with extended panels, the proportion of patients with a pathogenic or likely pathogenic variant was higher: 38–59% in the studies of the Kayvanpour meta-analysis and 42% in the French study of 95 adults, which included arrhythmia genes such as HCN4 and RYR2 [23]. In a series of 128 children with LVNC, of whom 75 underwent genetic testing, the yield was 9% [20]. As there are few LVNC-“exclusive” genes, the first step is an extended cardiomyopathy panel, with genes added according to the phenotype [11]. Clinical Cardiogenetics proposed a core panel of 45 genes (Table 5.4) and, in the absence of a panel, MYH7 sequencing as the first approach, as it is the most prevalent cause in adults and children; in severe adult or paediatric forms it recommended looking for a second molecular defect, given the frequency of multiple variants [1]. A French study of 4,900 index cases with DCM or hypertrabeculation found PRDM16 loss-of-function variants in 11 patients from nine families — including two whole-gene deletions and one multi-exon deletion — and the authors propose systematically including PRDM16, which not all panels analyse [42].

Table 5.4 — Panel of 45 nuclear genes proposed in Clinical Cardiogenetics (2020)

ACTC1, ACTN2, ANKRD1, BAG3, CALR3, CAV3, CRYAB, CSRP3, CTNNA3, DES, DSC2, DSG2, DSP, EMD, FHL1, GLA, JPH2, JUP, LAMA4, LAMP2, LMNA, LDB3, MIB1, MYBPC3, MYH6, MYH7, MYL2, MYL3, MYOZ2, MYPN, NEXN, PKP2, PLN, PRDM16, PRKAG2, RBM20, SCN5A, TAZ, TCAP, TMEM43, TNNC1, TNNI3, TNNT2, TPM1, TTN, TTR, VCL [1]

Several of these genes were subsequently classified by ClinGen as having limited or disputed evidence for the phenotypes concerned, and variants in genes with “moderate” or “limited” evidence should be interpreted with careful consideration of the genotype–phenotype correlation [22, 35].

The technical limitations of testing should be stated in the report. Deletion of exon 3 of RYR2 causes DCM or LVNC with CPVT and requires copy-number variation analysis (multiplex ligation-dependent probe amplification, MLPA, or sequencing-derived methods) [11, 43]. In TAFAZZIN, sequencing may not detect deletions of one or more exons or the entire gene, particularly in women, and a panel with deletion/duplication analysis is recommended; TAFAZZIN variants of uncertain significance can be clarified by the monolysocardiolipin/cardiolipin (MLCL/CL) ratio, which has high sensitivity and specificity [5]. Parental mosaicism with a low allele fraction may escape Sanger sequencing, whose detection limit is approximately 15–20% allele fraction, and require targeted deep sequencing (section 10) [50]. The report should state the gene, transcript with version (for example, NM_000257.4 for MYH7), Human Genome Variation Society (HGVS) nomenclature, zygosity, ACMG/Association for Molecular Pathology (AMP) classification and coverage limitations [5, 12]. Segregation testing of a familial VUS, particularly in parents or affected relatives, should be considered when it may enable confident interpretation (ESC, IIa-C) [3]. A negative test does not exclude genetic disease in the presence of a convincing familial phenotype: approximately half of familial forms remain without an identified cause [1].

5.5. Targeted non-genetic tests

When Barth syndrome is suspected, the MLCL/CL ratio in lymphocytes, fibroblasts, muscle or dried blood spot by tandem mass spectrometry is specific, sensitive and clinically available; 3-methylglutaconic aciduria suggests the diagnosis, but may be absent [5, 46]. When dystrophinopathy, laminopathy or myopathy is suspected, CK, electromyography and muscle biopsy guide the investigation; when a mitochondrial disorder is suspected, lactate, amino acids, organic acids and respiratory chain studies [1, 3]. In infants, the French PNDS indicates the fasting metabolic sampling described in section 4 and a full blood count to look for neutropenia [23]. In 1p36 deletion, chromosomal microarray defines whether PRDM16 is included, which alters the cardiac risk [41].

5.6. Integrated algorithm

1. Finding of excessive trabeculation: confirm the criteria by echocardiography (with contrast if necessary) and CMR; record function, dimensions, thickness, LGE and right ventricular involvement [3, 4].
2. Loading context: pregnancy, intense training, haemolytic anaemia, renal failure or hypertension — reassess after removal or stabilisation of the load [4].
3. Asymptomatic adult, normal function and morphology, normal ECG, no family history: do not perform genetic testing; no follow-up motivated solely by trabeculation [4, 11].
4. Associated cardiomyopathy (dilated, hypertrophic, restrictive, NDLVC, arrhythmogenic) or significant arrhythmia/conduction disturbance: classify according to the dominant phenotype, provide genetic counselling and use a cardiomyopathy panel; manage according to the phenotype-specific guideline [3, 11].
5. Child, fetus or extracardiac signs: metabolic, neuromuscular, mitochondrial and syndromic investigation; microarray, broad panel or exome; MLCL/CL ratio in boys [3–5].
6. Family: ECG and imaging in first-degree relatives; cascade testing if a pathogenic or likely pathogenic (P/LP) variant is present (section 8) [3].

The flowcharts in the 2019 HRS consensus (Figures 20 and 21 of the document) organise the same recommendations into decision trees, all with level of evidence B-NR (non-randomised data), and are reproduced in Table 5.5 for comparison with the algorithm above; unlike the latter, they start from LVNC as a disease and do not provide a branch for an incidental finding without cardiomyopathy [20].

Table 5.5 — Algorithms from the 2019 HRS consensus for LVNC

The legend to Figure 21 adds that aspirin is frequently administered in children. Source: Figures 20 and 21 and their respective legends [20].

6. Differential Diagnosis

The differential diagnosis has two aspects: distinguishing excessive trabeculation from physiological trabeculation and acquired forms, and, when disease is present, identifying the underlying cardiomyopathy or syndrome. In the normal heart, trabeculations are concentrated near the base of the mitral papillary muscles, there is no excessive segmental thickening, trabecular thickness does not exceed that of the compact layer, and the trabeculated area is smaller than in LVNC [1].

Table 6.1 — Differential diagnosis of the finding of excessive trabeculation

The distinction from HCM has a practical consequence that the Oxford Desk Reference highlights: management of stroke prophylaxis differs between HCM and LVNC, and sarcomeric variants can cause both phenotypes in different members of the same family [52]. The distinction from DCM with trabeculation no longer has its own therapeutic consequences: in a multicentre study of 1,160 patients, hypertrabeculation was not associated with embolic, arrhythmic or advanced heart failure events in any genotype group [16]. The French PNDS adds that noncompaction may coexist with DCM, HCM, RCM and Fabry disease, and that, in doubtful forms, both imaging techniques and the opinion of a specialised centre are recommended [23].

6.1. Syndromic and chromosomal forms

In the presence of dysmorphism, growth delay, intellectual disability, muscle weakness or other malformations, trabeculation should be interpreted as part of a syndrome (Table 6.2).

Table 6.2 — Syndromes and chromosomal abnormalities associated with LVNC/hypertrabeculation

Trabeculation has also been described in association with polycystic kidney disease, although it is not known whether the association reflects a genetic interaction or the increased preload of renal failure [4].

7. Treatments and Risk Management Strategies

7.1. Principle: treat the phenotype, not the trabeculation

Clinical Cardiogenetics summarises that treatment and follow-up depend largely on the coexisting cardiomyopathy, and that the general heart failure and ICD guidelines are applicable to the LVNC population [1]. The 2023 expert panel goes further: in adults with HCM or DCM in whom excessive trabeculation is also present, the extent of trabeculation has not been shown to alter management or prognosis, and it recommends following the specific guidelines for those diseases [4]. The 2023 ESC guidelines, by reclassifying LVNC as a trait, have no recommendations of their own for it: the indications for treatment, ICD, anticoagulation, exercise and reproduction are defined by the phenotype (DCM, NDLVC, HCM, RCM or arrhythmogenic right ventricular cardiomyopathy) [3]. No medicine is authorised for LVNC by the FDA or the EMA; the only medicine authorised for a specific genetic cause of hypertrabeculation is elamipretide for Barth syndrome, and only in the United States (7.6) [17, 19]. The French PNDS adopts the same rule: treatment is that of other DCMs, and there is no argument for treating patients with NCVG differently with regard to anticoagulation, arrhythmia prevention or the indication for an ICD [23].

7.2. Heart failure

When there is a dilated phenotype with symptomatic heart failure with reduced ejection fraction, Clinical Cardiogenetics indicates starting angiotensin-converting enzyme inhibitors and β-blockers according to the ESC heart failure guidelines [1]. Current pharmacological therapy for heart failure is not specific to LVNC and follows the heart failure guidelines in force. No LVNC-specific source defines doses; in the absence of a dose in the sources consulted, Table 7.1 presents the label dosage of each drug in adult heart failure with reduced ejection fraction, applied to the dilated phenotype by extrapolation. The French PNDS lists the same classes — angiotensin-converting enzyme inhibitors, gliflozins, mineralocorticoid receptor antagonists and β-blockers in left ventricular dysfunction, and diuretics in congestive heart failure — without giving doses, and considers that patients without ventricular dysfunction or symptoms do not necessarily need treatment, but should be monitored [23].

Table 7.1 — Label doses of drugs for heart failure with reduced ejection fraction

eGFR, estimated glomerular filtration rate; ACE inhibitor, angiotensin-converting enzyme inhibitor; NYHA, New York Heart Association. In children, use of these drugs without a paediatric label dose in heart failure is use outside the label conditions. The carvedilol label describes a trial in 161 children with chronic heart failure (mean age 6 years; 45% younger than 2 years), randomised to placebo or two dose levels, with no significant effect on clinical outcomes at 8 months [57]. In paediatric forms with systolic dysfunction, the PNDS applies DCM treatment; in forms with diastolic dysfunction, it recommends quantifying pulmonary pressures, whose elevation is a sign of severity, and treating as the restrictive forms, with a β-blocker (antiarrhythmic and heart rate-lowering) and diuretics according to symptoms, without giving doses; furosemide has a paediatric label dose (Table 7.1), and a β-blocker in children is use outside the label conditions [23, 57, 64].

Cardiac resynchronisation therapy, when indicated by the general criteria, has shown response rates above 50% in patients with LVNC [1]. Heart transplantation has been performed in patients with severe heart failure, and left ventricular restoration surgery has been described successfully in a single patient [1]. In the adult meta-analysis, the mean frequency of transplantation was 4% [10]. When pharmacological therapy is no longer sufficient, the PNDS provides, as in other cardiomyopathies, for pacemaker, resynchronisation, ICD, ventricular assist and transplantation, decided by multidisciplinary teams in specialised centres [23].

7.3. Prevention of sudden death and treatment of arrhythmias

Table 7.2 — ICD recommendations applicable to the patient with hypertrabeculation

LVEF, left ventricular ejection fraction; ARVC, arrhythmogenic right ventricular cardiomyopathy. Sources: [3, 20].

The French PNDS considers the arrhythmic risk of NCVG slightly higher than that of DCM, but with no study demonstrating benefit from a more aggressive approach, and applies the ICD indications for DCM, in line with the 2022 ESC guidelines on ventricular arrhythmias and the prevention of sudden death [23]. This position diverges from the 2019 HRS consensus, which retains LVNC-specific ICD recommendations (Table 7.2) [20].

ICD outcome data come from small series. In the Rotterdam cohort (67 patients), an ICD was indicated in 42% (21 for primary prevention, 7 for secondary prevention); over long-term follow-up, appropriate therapies occurred only in patients with a secondary prevention ICD (three), and inappropriate therapies affected 33% of primary prevention patients and 29% of secondary prevention patients [1]. In another study, among 12 patients with an ICD, 42% received appropriate therapy overall, 25% in primary prevention and 50% in secondary prevention; in long-term follow-up, appropriate and inappropriate shocks did not differ between HCM, DCM and LVNC [1]. In a prospective cohort of 44 patients with an ICD (12 for secondary prevention, 32 for primary prevention), eight received appropriate therapies over a median of 6.1 months and nine had inappropriate therapies; in a registry of 661 adults with LVNC and an ICD (78% for primary prevention), implantation complications were below 2% [20]. The adult meta-analysis found appropriate ICD therapy in 15% [10]. Catheter ablation was described in nine patients with VT or refractory frequent extrasystoles: the VT substrates coincided with the noncompacted segments, extrasystoles arose from the papillary muscles and basal septal region, and recurrence occurred in one patient after a median of 4 years, with improved function in half [20]. The expert panel notes that, after adjustment for dilatation, systolic dysfunction and fibrosis, trabeculation does not appear to confer additional arrhythmic risk [4]; in the multicentre DCM study, the predictors of major ventricular arrhythmias were genotype, ejection fraction and LGE, rather than hypertrabeculation [16].

Risk factors described in Clinical Cardiogenetics — in addition to resuscitated cardiac arrest and sustained VT/VF — include a family history of sudden death, unexplained syncope (particularly exertional), an abnormal blood pressure response to exercise, frequent ventricular extrasystoles, non-sustained VT on Holter monitoring and significant ventricular dysfunction [1]. In atrial fibrillation, catheter ablation is recommended after failure of or intolerance to a class I or III antiarrhythmic in patients with cardiomyopathy (I-B) [3].

7.4. Anticoagulation: the most relevant divergence

Table 7.3 — Positions on prevention of thromboembolism

Sources: [1, 3, 4, 16, 20, 23].

The HRS consensus based the anticoagulation recommendation on a cohort of 144 patients in which 22 strokes or peripheral embolisms occurred, 14 cardioembolic, related to atrial fibrillation or systolic dysfunction [20]. Subsequent evidence shifts the criterion to atrial fibrillation and ventricular function and takes weight away from trabeculation itself [4, 16]. In Barth syndrome, GeneReviews indicates aspirin to prevent thrombi in boys with severe cardiac dysfunction and/or marked noncompaction, at 5 mg/kg/day in children (no maximum dose defined in the source) and 81–325 mg/day in adults (maximum 325 mg/day) [5]. The use of aspirin for this purpose is not in the medicine's label; it is a GeneReviews recommendation for that syndrome. US professional labelling sets 50–325 mg/day after ischaemic stroke or transient ischaemic attack and 75–325 mg/day for prevention of recurrent myocardial infarction and in angina, a range that includes the adult dose by extrapolation; it has no paediatric antiplatelet dose and warns against use in children and adolescents with viral infections because of the risk of Reye syndrome, so the 5 mg/kg/day dose is use outside the label conditions [65].

7.5. Physical activity and sport

The HRS consensus recommends that patients with LVNC associated with arrhythmias, with or without systolic or diastolic dysfunction, avoid endurance exercise and competitive sport [20]. The 2023 ESC guidelines recommend individualised risk assessment for exercise prescription in all cardiomyopathies (I-C) and, in DCM and NDLVC, admit moderate- and high-intensity exercise in genotype-positive, phenotype-negative carriers (except pathogenic LMNA and TMEM43 variants) (IIa-C); they advise against high-intensity exercise with symptoms, LVEF ≤40%, exercise-induced arrhythmias or pathogenic LMNA or TMEM43 variants (III-C) [3]. The expert panel found no data justifying disqualification of athletes with isolated trabeculation and normal structure and function [4]. The French PNDS, relying on the North American (AHA/ACC 2015 and 2025) and European (ESC 2020) sports cardiology recommendations, encourages physical activity adapted to individual capacity and requires prior assessment with exercise testing and Holter for sports practice: with normal ejection fraction, no symptoms and normal results on these investigations, it imposes no restrictions and admits competitive sport; it advises against competition when there is impaired ejection fraction, arrhythmias, symptoms or fibrosis on CMR, and adjusts leisure sport case by case [23]. In left ventricular dysfunction, physical activity is not contraindicated, but should be adapted to the investigations; cardiac rehabilitation and moderate activity are recommended in stable heart failure [23]. The 2020 ESC guidelines, cited in the PNDS, provide for annual cardiological assessment of athletes with LVNC, including variant carriers without a phenotype [23]. The French position is more permissive than that of the HRS consensus, which recommends avoiding competitive and endurance sport in LVNC with arrhythmias, and coincides with that of ESC 2023 in advising against intense exercise with symptoms, dysfunction or arrhythmias [3, 20].

7.6. Barth syndrome: specific and supportive treatment

GeneReviews recommends standard heart failure treatment with careful fluid management, avoiding excessive diuresis and dehydration; angiotensin-converting enzyme inhibitors and β-blockers in outpatients; intravenous inotropes, including milrinone, in acute decompensation; and transplantation when heart failure is severe and intractable, weighed in the light of the natural improvement of ventricular function after the first year [5]. Transplant outcomes were similar to those of recipients without Barth syndrome, with less acute rejection [5]. Neutropenia is treated with granulocyte colony-stimulating factor (G-CSF, filgrastim) at the GeneReviews-recommended dose of 2–3 μg/kg/dose, at a frequency between twice weekly and alternate days, considering regular administration and antibiotic prophylaxis [5]. No label mentions Barth syndrome; for comparison, the labels for severe chronic neutropenia indicate 6 μg/kg SC twice daily in congenital neutropenia and 5 μg/kg once daily in idiopathic or cyclic neutropenia in the USA (Neupogen®), and 12 μg/kg/day in congenital and 5 μg/kg/day in idiopathic or cyclic neutropenia in the EU (Zarzio®), titrating to the lowest effective dose; the US label defines no maximum dose and mentions patients with congenital neutropenia who needed ≥100 μg/kg/day, and the EU SmPC states that 97% of complete responses occurred at ≤24 μg/kg/day, with no established long-term safety above that value [66, 67]; the GeneReviews dose is lower and less frequent, and its use in Barth syndrome lies outside the conditions of these labels. Prolonged fasting should be avoided or intravenous glucose given before procedures, and potassium monitored during potassium-containing fluids and episodes of diarrhoea [5].

Elamipretide (Forzinity®) is a tetrapeptide that binds mitochondrial cardiolipin. On 19 September 2025, it received FDA accelerated approval with priority review, as a new molecular entity, to improve muscle strength in adult and paediatric patients with Barth syndrome weighing ≥30 kg, on the basis of improved knee extension strength, an intermediate endpoint; continued approval may depend on confirmation of clinical benefit in a confirmatory trial [17, 18]. The labelled dosage is 40 mg subcutaneously (SC) once daily; in adults with an estimated glomerular filtration rate (eGFR) <30 mL/min who are not receiving dialysis, it is 20 mg once daily; no information is available for dialysis or paediatric patients with renal impairment [17]. The presentation is an injectable solution of 280 mg/3.5 mL (80 mg/mL) in a single-patient-use vial, with benzyl alcohol (20 mg/mL) as a preservative [17]. Injection-site reactions are the most frequent adverse effects (erythema in 12/12 patients receiving elamipretide and 3/12 receiving placebo during the controlled period; induration, pruritus and pain); severe hypersensitivity reactions and transient eosinophilia (peaking at approximately 90 days and resolving over 6–12 months) have been described, and the risk of benzyl alcohol toxicity contraindicates use in newborns [17]. In the European Union, elamipretide has orphan medicinal product designation for Barth syndrome (EU/3/21/2430, May 2021), but does not appear on the EMA list of authorised medicines or medicines under evaluation consulted in October 2026 [19].

Table 7.4 — Regulatory status of elamipretide

Sources: [17–19, 68].

Table 7.5 — Trials of elamipretide in Barth syndrome

Sources: [17, 68–70].

The limitations of these data are substantial. The controlled period was negative for both primary endpoints; the benefits come from an open-label extension without a control group, with eight patients remaining at the end, subject to training effects, expectancy effects and regression to the mean, and approval was based on a secondary intermediate endpoint [17, 69]. Reports of expanded-access use in newborns and critically ill patients do not constitute evidence of efficacy, and the label contraindicates use in newborns because of benzyl alcohol [17]. A registered intermediate-size expanded-access protocol exists (NCT04689360) [71].

7.7. Ongoing research and pipeline

For non-syndromic LVNC/hypertrabeculation, no therapeutic trials are registered on ClinicalTrials.gov; the registered studies are observational. The international consortium for multimodal phenotyping of adults with noncompaction (NCT04424030, Stanford, estimated 600 participants) has primary endpoints of the incidence of serious embolic and arrhythmic events over up to 4 years, with estimated primary completion in April 2027 [71]. The Chinese CAPTAIN study (NCT06504056, estimated 2,000 participants) aims to construct a prognostic prediction system [71]. For PRDM16-associated cardiomyopathy, PERM1 gene therapy was described in a preclinical model in a non-peer-reviewed preprint, without a clinical trial [72]. A search of registered trials does not exclude unregistered studies or studies in preparation.

7.8. General risk management

The ESC guidelines recommend lifestyle modification and treatment of intercurrent conditions to reduce the burden of atrial fibrillation and the severity of symptoms in patients with cardiomyopathy (I-B) [3]. Psychological support should be available within the multidisciplinary team, with attention to the impact of diagnosis, the ICD and exercise restrictions [3].

The French PNDS brings together lifestyle advice in a dedicated table: a balanced diet, avoiding salt and fluid overload, excessive alcohol consumption and large meals; information that β-blockers may disturb sexual life; compatibility with most occupations, except, with low ejection fraction, heart failure or arrhythmias (particularly ventricular), those involving intense physical exertion (athletes, armed forces, heavy load transport) and safety-critical posts (pilots, drivers) [23]. Private driving is permitted and NCVG does not, in itself, contraindicate air travel; a cardiology consultation is recommended before travel, to assess individual risk, together with vaccination against influenza and pneumococcus [23]. The announcement of the diagnosis should explain the nature of the disease, its genetic implications, the treatment options, the risks of complications and the surveillance required, and therapeutic education of the patient and family includes recognition of the signs of complications and adherence to treatment and follow-up [23]. The team envisaged includes cardiologists, geneticists, paediatricians, internal medicine physicians, general practitioners, school doctors, sports physicians and obstetricians [23].

8. Reproductive Options and Family Counselling

8.1. Inheritance patterns and recurrence risk

In isolated familial forms, inheritance is predominantly autosomal dominant, with a 50% risk for each child of a carrier [1, 20]. Expression in relatives does not necessarily reproduce the proband’s phenotype: in the Dutch families, screening identified 73 relatives with LVNC and 34 with DCM or HCM without LVNC, and the relatives’ phenotype could be predicted from the proband’s phenotype and genotype (isolated LVNC with variants in the head of MYH7 and isolated LVNC in relatives; LVNC with dilatation and DCM in relatives; LVNC with hypertrophy and HCM in relatives) [36]. In X-linked forms (Barth syndrome), the mother of an affected boy with more than one affected relative is an obligate heterozygote; in a simplex case, the variant may be de novo, maternally inherited or the result of maternal gonadal mosaicism — in one laboratory, the proband’s variant was not found in the leucocyte DNA of 5 of 42 mothers [5]. A heterozygous mother transmits the variant in 50% of pregnancies; boys who inherit it are affected and heterozygous girls usually do not manifest the disease, because of preferential inactivation of the X chromosome carrying the variant; an affected man transmits it to all his daughters and none of his sons [5]. In autosomal recessive and mitochondrial forms, counselling follows the rules of those inheritance patterns [11]. Parental somatic or gonadal mosaicism changes the recurrence risk of apparently de novo variants (section 10) [5, 50].

8.2. Clinical and genetic screening of relatives

The familial occurrence of isolated adult forms has been estimated at between 18% and 71%, and, because most affected relatives are asymptomatic, family history alone is insufficient: Clinical Cardiogenetics recommends cardiac assessment of all adult relatives regardless of clinical history [1]. In a study of 194 relatives of 50 probands, 64% of families had familial cardiomyopathy, including HCM and DCM, and 8 of 17 asymptomatic carriers had no phenotype [20]. In the Dutch families, screening had a higher yield when a variant had been identified, and 37% of relatives carrying a variant had no phenotype [36].

Table 8.1 — Family screening recommendations

Sources: [3, 11, 20, 23].

For carriers of P/LP variants or in families with demonstrated disease, the ESC guidelines propose ECG and echocardiography every 1–3 years up to 60 years of age and every 3–5 years thereafter, from childhood; penetrance in familial cardiomyopathies is about 5% in the first decade, 10–20% per decade between the second and seventh, and reaches 70–90% at 70 years, with a delay of about 10 years in women [3]. In informative families in which the proband is the only affected member and genetic testing was negative, ending periodic surveillance may be considered in first-degree relatives aged ≥50 years with normal investigations [3]. These recommendations apply to families with cardiomyopathy; the penetrance of the same variants found outside that context is probably much lower [3]. The French PNDS estimates 10–40% familial forms and justifies continued surveillance of relatives with a normal first examination by the possible overlap with DCM or HCM in the same family and by the rare forms of late cardiac expression; it differs from ESC 2023 in starting echocardiographic screening from birth, without defined intervals, and in providing neither for discharge of non-carrier relatives nor for ending surveillance at 50 years [3, 23].

The Clinical Cardiogenetics flowchart (Figure 7.5) stratifies follow-up according to the proband’s result: with a class 5 variant, predictive testing and, if phenotype-negative, follow-up every 1–2 years; with a class 4 variant, testing combined with cardiac screening and follow-up every 2 years in phenotype-negative carriers and every 5 years in phenotype-negative non-carriers; with a VUS (class 3), cardiac screening, testing only if a phenotype is present and follow-up every 4 years if negative; without a variant, cardiac screening with follow-up every 4 years [1]. These intervals are a 2020 proposal from a Dutch centre and do not coincide with those of the 2023 ESC guidelines, which should take precedence in the European setting. Screening of relatives may reveal minimal abnormalities that do not fulfil criteria and are difficult to distinguish from physiological trabeculation; their significance for subsequent progression is not established [1].

8.3. Reproductive options

The ESC guidelines recommend counselling men and women about the risk of transmission before conception (I-C) and state that discussion of reproductive genetic testing options with a trained professional should be considered in all families with a genetic diagnosis (IIa-C); if prenatal diagnosis is chosen, it should be performed early in pregnancy (I-C) [3]. Options include prenatal diagnosis by chorionic villus sampling at 10–14 weeks (procedure-related fetal loss of approximately 0.2%) or amniocentesis after 15 weeks (approximately 0.1%), non-invasive prenatal testing for monogenic disease, which remains of limited availability, and preimplantation genetic diagnosis, with a success rate of 25–30% per embryo transfer and the risks inherent to in vitro fertilisation [3]. At a Dutch preimplantation genetic diagnosis centre, only one of the 45 referred couples with inherited cardiac disease had noncompaction cardiomyopathy [73]. In Barth syndrome, prenatal and preimplantation diagnosis is possible when the familial variant is known; because of the risks of growth restriction, oligohydramnios, ventricular dysfunction and fetal hydrops, pregnancies involving affected male fetuses should be followed by maternal–fetal medicine, and delivery should take place in a tertiary centre with intensive care [5].

8.4. Pregnancy in women with hypertrabeculation

Pregnancy itself may induce reversible trabeculation in about a quarter of women with a normal heart [28]. In women with LVNC, maternal risk depends on the associated cardiomyopathy or congenital heart disease: according to the modified World Health Organization (mWHO) classification cited in Clinical Cardiogenetics, women in class IV — previous peripartum cardiomyopathy with residual dysfunction, ejection fraction <30% or New York Heart Association (NYHA) functional class III–IV, severe mitral or aortic stenosis — have an event rate of 40–100%, which contraindicates pregnancy, and those in classes II–III (for example, HCM, moderate valve disease, atrioventricular septal defect, repaired coarctation) have a rate of 10–19% [1]. The 2023 ESC guidelines recommend preconception assessment and counselling with the mWHO classification (I-C), contraceptive counselling (I-C), prior review of medication (I-C), vaginal delivery in most women, except for obstetric indications, severe heart failure (ejection fraction <30% or NYHA III–IV), severe obstruction or labour on oral anticoagulation (I-C), continuation of β-blockers with monitoring of fetal growth (IIa-C), anticoagulation with low-molecular-weight heparin or vitamin K antagonists according to the stage of pregnancy in atrial fibrillation (I-C), and genetic counselling and testing in peripartum cardiomyopathy (IIa-C) [3]. The French PNDS applies to women with NCVG the approach used for other cardiomyopathies: contraception discussed and adapted to the risk–benefit ratio, preconception assessment before stopping contraception, pregnancy planned with a multidisciplinary cardiology and obstetrics team, and contraindication decided case by case, particularly in the most severe forms or those at high arrhythmic risk [23]. The particular feature for NCVG is the recommendation of fetal echocardiography to detect a form of prenatal onset; regular cardiological surveillance is reserved for women with left ventricular dysfunction, and in the absence of dysfunction conventional obstetric surveillance is sufficient [23].

9. Special Considerations: Population Diversity and Epidemiology

9.1. What prevalence estimates measure

Orphanet classifies the point prevalence of LVNC as “unknown” [2]. Published estimates are not population prevalences of the disease: they measure the frequency of an imaging criterion in a selected population and vary according to the technique, criterion and population (Table 9.1).

Table 9.1 — Frequency of LVNC/excessive trabeculation criteria by population

The 2019 HRS consensus reports an estimate of about 1 case per 7,000 live births and genetic inheritance in at least 30–50% of patients [20]. These estimates, like the series of referred patients, should not be read as national prevalence: they depend on selection and criteria [4]. The incidence of primary cardiomyopathies in children is about 1 per 100,000 person-years in population-based studies from the United States, Finland and Australia, with higher values in boys, Black children and Aboriginal Australian children [14]. The paediatric incidence value of the French PNDS (greater than 1 per 100,000 children) is difficult to reconcile with this overall incidence of primary cardiomyopathies in children, of which LVNC represents about 5–9%; the document indicates neither the source population nor the denominator [14, 23]. According to the same PNDS, the isolated form predominates in adults and the forms associated with other heart diseases, neuromuscular diseases, syndromes or chromosomal disorders predominate in children [23].

9.2. Sex and ethnicity

Male predominance (ratio close to 2:1; two thirds of patients in the adult meta-analysis) is consistent across clinical series [1, 10]. The French PNDS gives a lower male:female ratio, of 1.5:1, and shows, in the French registry, diagnoses at all ages, from the fetus to the elderly [23]. In 1p36 deletion and in PRDM16 loss-of-function variants, the effect of sex is reversed: 34.5% of women with a PRDM16 deletion developed cardiomyopathy, versus 16.7% of men, and in the French series the median age at diagnosis was 18.5 years in women and 49 years in men [41, 42]. Trabeculation appears more frequent in people of African ancestry: in MESA, African American participants had a greater fractal dimension than White participants; in pregnancy, African American women were three times more likely to develop trabeculation than Caucasian women; and four of the five healthy controls who met echocardiographic criteria in a study of systolic dysfunction were Black [4, 20]. Clinical Cardiogenetics warns that further studies are needed to confirm whether excessive trabeculation is more prevalent in specific ethnic groups [1]. Uncritical application of thresholds derived from European populations may thus increase overdiagnosis in people of African ancestry.

9.3. Regional data and founder populations

The large genetic series come from the Netherlands (registry of 327 probands from four centres), Germany (registry of 95 patients with exome sequencing), Switzerland and Germany, France, Poland, Italy and the United States, and the population imaging cohorts from the United Kingdom (UK Biobank) and the United States (MESA) [4, 9, 33, 34]. The 2026 multicentre study on hypertrabeculation in DCM brought together mainly Spanish centres of the CIBER Cardiovascular network and the Amsterdam centre of ERN GUARD-Heart [16]. In Japan, 79 patients with familial LVNC showed a predominance of autosomal dominant transmission, and the DTNA variant underlying OMIM entry 604169 was described in a Japanese family [7, 14]. Barth syndrome has an estimated prevalence, by Bayesian analysis, of 1 per million boys, with 230–250 patients identified worldwide [5]. France has its own data: the first multicentre registry of the French Society of Cardiology (105 adults with isolated LVNC, published in 2011), a prospective multicentre cohort of 98 adults with LVNC compared with 65 patients with DCM and the genetic study of 95 unrelated adults from that cohort [23]. No founder variants associated with LVNC, and no specific data from Portugal or Brazil, were identified in the sources consulted; this absence in the sources consulted does not demonstrate that no published studies exist.

9.4. Resources

The European Reference Network for rare, low-prevalence and complex heart diseases (ERN GUARD-Heart) covers inherited cardiomyopathies and paediatric cardiogenetics [24]. Patient associations include the Barth Syndrome Foundation (United States), Barth Syndrome Trust (United Kingdom), Children’s Cardiomyopathy Foundation (United States) and Cardiomyopathy UK [74–77].

In France, the organisation rests on the reference centre for inherited or rare cardiomyopathies and cardiac rhythm disorders, coordinated at the Pitié-Salpêtrière (Paris), with constituent sites in Paris (Ambroise-Paré, Bichat, Hôpital Européen Georges-Pompidou, Henri-Mondor, Necker), Nantes, Bordeaux, Toulouse, Marseille (La Timone) and Rennes, and on the competence centres of the national CARDIOGEN health filière, whose directory lists the contacts [23]. The filière maintains a national psychological resource centre, a public information page on NCVG and the 2024 expert consensus on counselling and genetic testing [23]. The patient associations indicated are the Ligue contre la Cardiomyopathie, dedicated to all types of cardiomyopathy, and APODEC, the association of defibrillator carriers [23]. The PNDS serves as a reference for the attending physician for the care protocol drawn up with the medical adviser of the Caisse d'assurance maladie, when exemption from the patient's co-payment (ticket modérateur) is requested as an off-list long-term condition [23].

10. Illustrative Clinical Case

10.1. Recurrent fetal hydrops caused by LVNC associated with low-fraction maternal mosaicism for a MYH7 variant (Kawamura et al., 2023)

A 35-year-old Japanese woman, gravida 4, para 2, without a relevant personal or family history, was followed at a university hospital in Fukui [50]. Her first pregnancy ended in a first-trimester miscarriage and her second, at 28 years, in a healthy term newborn girl [50]. During her third pregnancy, at 33 years, she was referred at 23+5 weeks because of fetal hydrops: generalised oedema, ascites, bilateral pleural effusions, cardiomegaly (cardiothoracic area ratio 56.7%), severe tricuspid regurgitation, moderate mitral regurgitation and biventricular hypokinesia, with myocardial performance indices of 0.92 (left) and 0.85 (right) [50]. Fetal echocardiography showed noncompacted layers in both ventricles and an extensive trabeculated layer with deep intertrabecular recesses filled from the left ventricular cavity — hydrops of cardiac origin caused by LVNC [50]. At 31+0 weeks, after spontaneous labour and vaginal delivery agreed jointly with the parents because of the fatal prognosis, a boy weighing 2,440 g was born, with Apgar scores of 1 and 1, and died shortly after birth despite resuscitation; the parents declined autopsy and genetic testing [50].

During the fourth pregnancy, the fetus was monitored every 2–3 weeks from 13 weeks. At 21+1 weeks there were no abnormalities; at 23+1 weeks, reduced ventricular contractility, moderate tricuspid and mitral regurgitation and myocardial imaging appearances of LVNC similar to those of the previous fetus emerged; at 25+4 weeks, signs of hydrops developed and worsened [50]. Suspected familial LVNC led to a shared decision to perform a caesarean section at 32+0 weeks to prevent fetal death, after the appearance of absent or reversed diastolic flow in the umbilical artery and an oedematous placenta; a boy weighing 2,680 g was born, with Apgar scores of 1 and 1, and died shortly afterwards [50]. At autopsy, there was generalised oedema, effusions, a markedly dilated right ventricle and a hypertrophic left ventricle; histology showed a bilayered myocardium with a prominent trabeculated layer and endocardial fibroelastic proliferation throughout the left ventricle [50].

Next-generation sequencing of 182 cardiomyopathy and channelopathy genes in the newborn’s blood identified a novel heterozygous missense variant in MYH7, NM_000257.3:c.2729A>T, p.(Lys910Ile), absent from ClinVar, HGMD, dbSNP, gnomAD and the Japanese jMorp panel [50]. In the current MANE Select transcript, the variant corresponds to NM_000257.4:c.2729A>T, NP_000248.2:p.(Lys910Ile), NC_000014.9:g.23424100T>A (GRCh38), according to VariantValidator [12]. Sanger sequencing of the parents did not detect the variant in the father and showed only a minimal change of uncertain significance in the mother’s chromatogram; targeted deep sequencing found the variant in 6% of maternal reads (807/14 266) and 0% of paternal reads (0/1 658) [50]. The parents had normal echocardiograms, the mother had no muscle weakness and maternal CK was normal [50].

Figures from the original article (open access, CC BY licence): Figure 1 — fetal echocardiography from the two pregnancies (asterisk on thickened myocardium, arrows indicating flow into the recesses), fetal oedema and histology with compact and noncompacted layers and endocardial fibrosis (https://pmc.ncbi.nlm.nih.gov/articles/PMC10285293/figure/F1/); Figure 2 — proband sequencing and pedigree (https://pmc.ncbi.nlm.nih.gov/articles/PMC10285293/figure/F2/); Figure 3 — Sanger chromatograms of the trio, with the minimal maternal change indicated (https://pmc.ncbi.nlm.nih.gov/articles/PMC10285293/figure/F3/); Figure 4 — workflow from the previous delivery to parental sequencing (https://pmc.ncbi.nlm.nih.gov/articles/PMC10285293/figure/F4/) [50].

10.2. Commentary

The case illustrates five points in this article. First, the fetal form can be biventricular and severe, unlike the incidental finding in adults, and MYH7 variants are among the most frequent causes of fetal-onset LVNC (7 of 15 variants in a series of 33 cases cited by the authors) [50]. Second, the same gene associated with the lowest risk of adverse events in adult series can cause the most severe form when presentation is fetal; genotype does not replace phenotype in estimating prognosis [9, 50]. Third, an apparently de novo variant may result from low-fraction parental mosaicism below the detection limit of Sanger sequencing (15–20% allele fraction), which changes recurrence risk and justifies targeted deep sequencing of the parents [50]. Fourth, maternal assessment was limited to peripheral blood; gonadal mosaicism could not be measured without an invasive procedure, and genes outside the 182-gene panel were not analysed [50]. Fifth, the authors considered the variant pathogenic on the basis of in silico predictors (GADD and PolyPhen-2 scores), its absence from population databases and recurrence of the phenotype in the two brothers, without describing formal application of the ACMG/AMP criteria; in a clinical report, classification should result from integration of all criteria rather than computational predictors alone [50]. The obstetric management described (serial surveillance, with timing and mode of delivery decided with the parents) responds to the specific situation and does not constitute a routine recommendation.

11. Prognosis and Long-Term Follow-up

11.1. From initial severity to heterogeneous prognosis

The earliest series described a severe prognosis, but detection of asymptomatic patients through new imaging techniques showed that those observations reflected selection of the most severe cases; prognosis is as variable as in other cardiomyopathies, with gradual improvement in function even in some early-childhood presentations and progression to transplantation in others [1]. NYHA functional class III or higher and cardiovascular complications predict a poor prognosis, whereas age does not do so in children [1]. Patients with LVNC and DCM have a worse prognosis than those with HCM or sporadic LVNC, and genotype-negative patients without overt cardiomyopathy have an excellent prognosis [1].

Table 11.1 — Determinants of prognosis

In children, data from the American PCMR registry show that prognosis is determined by the associated phenotype: in 155 children with LVNC, the 5-year mortality rate was 6% (95% CI 2–22) in isolated LVNC, 22% (13–37) with a dilated phenotype, 25% (10–53) with a hypertrophic phenotype and 9% (1–49) with an indeterminate phenotype; there were 16 transplants in the dilated group and none in the isolated group [14]. In the Texas Children’s Hospital series (242 children with isolated LVNC), 12.8% died and 5.4% underwent transplantation, and 62% presented with or developed dysfunction [48]. Among 36 children diagnosed predominantly during the first year, 14% died during the study, and transient recovery followed by late deterioration (median 6.3 years) may explain some adult presentations [47]. In Barth syndrome, function improves in many patients by mid-childhood, but the registry showed a mean decrease of 0.6 in the ejection fraction z-score for each 5 years of age, and adolescents and adults tend towards a restrictive phenotype [5].

In the adult meta-analysis (2,271 patients), systemic thromboembolic events occurred in 9%, transplantation in 4% and appropriate ICD therapy in 15%; these frequencies come from series of referred patients with heterogeneous follow-up and do not represent the risk of an incidental finding [10]. Likewise, the absence of serious events in low-risk subgroups over approximately 5 years of follow-up does not equate to a zero lifetime risk [15].

The French PNDS interprets the historical evolution in the same way: in the first French registry (2010–2011), as in other older series, the overall prognosis was poor, whereas, in the second prospective registry, complications were rarer and concentrated in patients with left ventricular dysfunction; at equal ejection fraction, however, heart failure admission was more frequent in LVNC than in DCM, with similar results in a North American long-term survival study [23]. A systematic review with meta-analysis of 28 observational studies (2,501 patients) cited in the PNDS showed considerable variability in manifestations and long-term prognosis [23]. The conclusion of the PNDS — excellent prognosis with preserved function and higher risk than DCM with dysfunction — rests on cohorts of referred patients with follow-up of a few years, and should not be read as a zero lifetime risk with preserved function [23].

11.2. Follow-up

Table 11.2 — Recommended follow-up

Intervals diverge between documents: the 2023 panel dispenses with follow-up in the adult with an incidental finding, normal function, morphology and ECG and no family history, whereas the French PNDS provides for regular surveillance of all asymptomatic patients; for disease with dysfunction, the PNDS sets an annual minimum, tighter than the 1–2-year interval of ESC 2023 for stable cardiomyopathy, and in relatives it starts echocardiography at birth, without the 1–3- and 3–5-year intervals of the ESC [3, 4, 23].

Transition from paediatric to adult medicine should preserve genetic information, serial imaging results and the family screening plan. Pregnancy, the start of intense training and intercurrent episodes involving volume overload are times when the trabecular pattern may change without signifying disease progression [4, 28].

12. Summary and Final Considerations

Over three decades, left ventricular noncompaction has moved from a primary genetic cardiomyopathy (AHA 2006) to a phenotypic trait (ESC 2023), with a European recommendation to use the term “hypertrabeculation” [3, 8]. This change rests on three observations: developmental biology does not confirm embryonic “compaction” whose arrest would explain the phenotype; imaging criteria are fulfilled by a substantial fraction of the healthy population, athletes and pregnant women; and trabeculation adds no prognostic value beyond ventricular function, LGE and genotype [4, 13, 15, 16]. Divergence between documents persists: the 2019 HRS consensus retains LVNC-specific ICD and anticoagulation recommendations, and the 2019 AHA paediatric statement retains the category “LVNC cardiomyopathy” [14, 20].

For the geneticist, the central question is not whether the ventricle “is noncompacted”, but which cardiac and extracardiac phenotype accompanies trabeculation. Genetic testing should not be performed in asymptomatic adults with normal function, a normal ECG and no family history [11]. When associated cardiomyopathy, dysfunction, arrhythmia, a family history, paediatric age or syndromic signs are present, a cardiomyopathy panel identifies a cause in 17–41% of cases, particularly MYH7, MYBPC3 and TTN, and less frequently ACTC1, TNNT2, TPM1, LMNA, LDB3, MIB1, PRDM16, RYR2 or HCN4; truncating variants in MYH7, ACTN2 and PRDM16 are those most specifically associated with trabeculation [9, 11, 34]. In children, Barth syndrome, dystrophinopathies, mitochondrial diseases and chromosomal syndromes must be actively sought [1, 4, 5].

The OMIM entries sometimes associated with the designation (604169, DTNA; 601493, LDB3) correspond to associations with limited or disputed evidence, and not to the most frequent genes; the genes often cited (TTN, MYH7, LMNA) are, in fact, DCM and HCM genes in which trabeculation appears as a phenotypic expression [6, 10, 35]. Treatment is that of the associated cardiomyopathy; anticoagulation should be based on atrial fibrillation, previous embolism and ventricular function, and the 2026 evidence does not support prophylactic anticoagulation for trabeculation in DCM in sinus rhythm [3, 16, 20]. The only targeted therapy authorised for a genetic cause of hypertrabeculation is elamipretide for Barth syndrome, with accelerated approval in the United States based on an intermediate endpoint, a confirmatory trial under way and no European authorisation [17, 19, 68]. Family counselling follows the cardiomyopathy rule: cascade testing when there is a P/LP variant, clinical screening of first-degree relatives when there is not, and attention to parental mosaicism when the variant appears de novo [3, 50]. The 2026 French PNDS, which continues to treat noncompaction as a genetic disease, converges with these positions on treatment (that of DCM, including ICD and anticoagulation) and diverges on the breadth of investigation and follow-up: CMR in all suspected cases, genetic testing of the index case, echocardiographic screening of relatives from birth and regular surveillance of asymptomatic individuals [23].

13. References

1. Hoedemaekers YM, Germans T. Left ventricular noncompaction. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. Chapter 7, p. 115–138. doi:10.1007/978-3-030-45457-9_7
2. Orphanet. Left ventricular noncompaction. ORPHA:54260 [Internet]. Definition, synonyms, ICD-10 (I42.8) and ICD-11 (BC44) mappings, OMIM mappings and prevalence class (“unknown”) accessed in the Orphacodes API and Orphadata prevalence file on 7 October 2026. Disease entry
3. Arbelo E, Protonotarios A, Gimeno JR, Arbustini E, Barriales-Villa R, Basso C, et al. 2023 ESC Guidelines for the management of cardiomyopathies. Eur Heart J. 2023;44(37):3503–3626. doi:10.1093/eurheartj/ehad194
4. Petersen SE, Jensen B, Aung N, Friedrich MG, McMahon CJ, Mohiddin SA, et al. Excessive trabeculation of the left ventricle: JACC: Cardiovascular Imaging Expert Panel Paper. JACC Cardiovasc Imaging. 2023;16(3):408–425. doi:10.1016/j.jcmg.2022.12.026
5. Ferreira C, Pierre G, Thompson R, Vernon H. Barth syndrome. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle: University of Washington; 1993–2026. Published on 9 October 2014; updated on 8 January 2026. NBK247162
6. Online Mendelian Inheritance in Man (OMIM). Left ventricular noncompaction — phenotypic series PS604169; entries 604169, 601493, 601494, 609470, 611878, 613424, 613426, 615092, 615373 and 615396. Baltimore: Johns Hopkins University. Phenotypic series and entries 604169, 601493 and 609470 accessed on 7 October 2026. PS604169
7. Ichida F, Tsubata S, Bowles KR, Haneda N, Uese K, Miyawaki T, et al. Novel gene mutations in patients with left ventricular noncompaction or Barth syndrome. Circulation. 2001;103(9):1256–1263. doi:10.1161/01.cir.103.9.1256
8. Maron BJ, Towbin JA, Thiene G, Antzelevitch C, Corrado D, Arnett D, et al. Contemporary definitions and classification of the cardiomyopathies: an American Heart Association Scientific Statement. Circulation. 2006;113(14):1807–1816. doi:10.1161/CIRCULATIONAHA.106.174287
9. van Waning JI, Caliskan K, Hoedemaekers YM, van Spaendonck-Zwarts KY, Baas AF, Boekholdt SM, et al. Genetics, clinical features, and long-term outcome of noncompaction cardiomyopathy. J Am Coll Cardiol. 2018;71(7):711–722. doi:10.1016/j.jacc.2017.12.019
10. Kayvanpour E, Sedaghat-Hamedani F, Gi WT, Tugrul OF, Amr A, Haas J, et al. Clinical and genetic insights into non-compaction: a meta-analysis and systematic review on 7598 individuals. Clin Res Cardiol. 2019;108(11):1297–1308. doi:10.1007/s00392-019-01465-3
11. Wilde AAM, Semsarian C, Márquez MF, Shamloo AS, Ackerman MJ, Ashley EA, et al. European Heart Rhythm Association (EHRA)/Heart Rhythm Society (HRS)/Asia Pacific Heart Rhythm Society (APHRS)/Latin American Heart Rhythm Society (LAHRS) Expert Consensus Statement on the state of genetic testing for cardiac diseases. Europace. 2022;24(8):1307–1367. doi:10.1093/europace/euac030
12. VariantValidator (version 4.0.1; vvdb_2025_3): MANE Select transcripts by gene and validation of NM_000257.3:c.2729A>T. Accessed on 7 October 2026. rest.variantvalidator.org
13. Ross SB, Jones K, Blanch B, Puranik R, McGeechan K, Barratt A, et al. A systematic review and meta-analysis of the prevalence of left ventricular non-compaction in adults. Eur Heart J. 2020;41(14):1428–1436. doi:10.1093/eurheartj/ehz317
14. Lipshultz SE, Law YM, Asante-Korang A, Austin ED, Dipchand AI, Everitt MD, et al. Cardiomyopathy in children: classification and diagnosis: a scientific statement from the American Heart Association. Circulation. 2019;140(1):e9–e68. doi:10.1161/CIR.0000000000000682
15. Grigoratos C, Barison A, Ivanov A, Andreini D, Amzulescu MS, Mazurkiewicz L, et al. Meta-analysis of the prognostic role of late gadolinium enhancement and global systolic impairment in left ventricular noncompaction. JACC Cardiovasc Imaging. 2019;12(11 Pt 1):2141–2151. doi:10.1016/j.jcmg.2018.12.029
16. Mora-Ayestarán N, Ramos-Lopez N, Ochoa JP, Ibargoyen AY, Claver-Garrido E, Gómez-González C, et al. Left ventricular hypertrabeculation and prognosis in dilated cardiomyopathy. Circulation. 2026;154(5):440–453. doi:10.1161/CIRCULATIONAHA.125.078536
17. Stealth BioTherapeutics Inc. FORZINITY™ (elamipretide) injection, for subcutaneous use. Prescribing information (U.S. label), version dated 10 December 2025; NDA 215244. DailyMed setid 146bf34c-76f2-48db-ac07-fb29cce2cd75
18. U.S. Food and Drug Administration. Drugs@FDA: NDA 215244 — FORZINITY (elamipretide); original submission approved on 19 September 2025 (priority review; new molecular entity). Accessed via openFDA on 7 October 2026. Drugs@FDA
19. European Medicines Agency. Orphan medicinal product designation EU/3/21/2430 — elamipretide, treatment of Barth syndrome (designation dated 20 May 2021); orphan designation and medicines (EPAR) registers accessed on 7 October 2026. Medicines data
20. Towbin JA, McKenna WJ, Abrams DJ, Ackerman MJ, Calkins H, Darrieux FCC, et al. 2019 HRS expert consensus statement on evaluation, risk stratification, and management of arrhythmogenic cardiomyopathy. Heart Rhythm. 2019;16(11):e301–e372. doi:10.1016/j.hrthm.2019.05.007
21. Hershberger RE, Jordan E. Dilated cardiomyopathy overview. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle: University of Washington; 1993–2026. Published on 27 July 2007; updated on 29 July 2021; reviewed on 12 December 2024. NBK1309
22. Cirino AL, Channaoui N, Ho C. Nonsyndromic hypertrophic cardiomyopathy overview. In: Adam MP, et al., editors. GeneReviews® [Internet]. Seattle: University of Washington; 1993–2026. Published on 5 August 2008; updated on 6 March 2025. NBK1768
23. Haute Autorité de Santé; Centre de référence des maladies cardiaques héréditaires ou rares (cardiomyopathies et troubles du rythme), Hôpital de La Timone (AP-HM, Marseille) and Hôpital de la Pitié-Salpêtrière (AP-HP, Paris); filière CARDIOGEN. Protocole National de Diagnostic et de Soins (PNDS): Non-compaction du ventricule gauche (NCVG) (hypertrabéculation ventriculaire gauche). Version du 8 septembre 2025. Saint-Denis La Plaine: HAS; 6 fév 2026 [accessed 2026-11-02]. Available from: https://www.has-sante.fr/jcms/p_3858430
24. European Reference Network for rare, low prevalence and complex diseases of the heart (ERN GUARD-Heart). guardheart.ern-net.eu
25. Feldt RH, Rahimtoola SH, Davis GD, Swan HJ, Titus JL. Anomalous ventricular myocardial patterns in a child with complex congenital heart disease. Am J Cardiol. 1969;23(5):732–734. doi:10.1016/0002-9149(69)90037-x
26. Chin TK, Perloff JK, Williams RG, Jue K, Mohrmann R. Isolated noncompaction of left ventricular myocardium: a study of eight cases. Circulation. 1990;82(2):507–513. doi:10.1161/01.cir.82.2.507
27. Jenni R, Oechslin E, Schneider J, Attenhofer Jost C, Kaufmann PA. Echocardiographic and pathoanatomical characteristics of isolated left ventricular non-compaction: a step towards classification as a distinct cardiomyopathy. Heart. 2001;86(6):666–671. doi:10.1136/heart.86.6.666
28. Gati S, Papadakis M, Papamichael ND, Zaidi A, Sheikh N, Reed M, et al. Reversible de novo left ventricular trabeculations in pregnant women: implications for the diagnosis of left ventricular noncompaction in low-risk populations. Circulation. 2014;130(6):475–483. doi:10.1161/CIRCULATIONAHA.114.008554
29. Hershberger RE, Morales A, Cowan J. Is left ventricular noncompaction a trait, phenotype, or disease? The evidence points to phenotype. Circ Cardiovasc Genet. 2017;10(6):e001968. doi:10.1161/CIRCGENETICS.117.001968
30. McGurk KA, Halliday BP, O'Regan DP. Left ventricular noncompaction: time to retire a flawed diagnosis. J Am Coll Cardiol. 2026;87(25):3651–3653. doi:10.1016/j.jacc.2026.03.052
31. van der Zwaag PA, van den Berg MP. Introduction to hereditary cardiomyopathies. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. Chapter 3, p. 53–61 (section 3.5, p. 59). doi:10.1007/978-3-030-45457-9_3
32. Klaassen S, Probst S, Oechslin E, Gerull B, Krings G, Schuler P, et al. Mutations in sarcomere protein genes in left ventricular noncompaction. Circulation. 2008;117(22):2893–2901. doi:10.1161/CIRCULATIONAHA.107.746164
33. Sedaghat-Hamedani F, Haas J, Zhu F, Geier C, Kayvanpour E, Liss M, et al. Clinical genetics and outcome of left ventricular non-compaction cardiomyopathy. Eur Heart J. 2017;38(46):3449–3460. doi:10.1093/eurheartj/ehx545
34. Mazzarotto F, Hawley MH, Beltrami M, Beekman L, de Marvao A, McGurk KA, et al. Systematic large-scale assessment of the genetic architecture of left ventricular noncompaction reveals diverse etiologies. Genet Med. 2021;23(5):856–864. doi:10.1038/s41436-020-01049-x
35. Clinical Genome Resource (ClinGen). Gene-Disease Validity Curations — export file created on 7 Oct 2026 (MYH7, MYBPC3, TTN, ACTC1, TNNT2, TPM1, LMNA, LDB3, DTNA, MIB1, PRDM16, ACTN2, TAFAZZIN). search.clinicalgenome.org
36. van Waning JI, Caliskan K, Michels M, Schinkel AFL, Hirsch A, Dalinghaus M, et al. Cardiac phenotypes, genetics, and risks in familial noncompaction cardiomyopathy. J Am Coll Cardiol. 2019;73(13):1601–1611. doi:10.1016/j.jacc.2018.12.085
37. Xing Y, Ichida F, Matsuoka T, Isobe T, Ikemoto Y, Higaki T, et al. Genetic analysis in patients with left ventricular noncompaction and evidence for genetic heterogeneity. Mol Genet Metab. 2006;88(1):71–77. doi:10.1016/j.ymgme.2005.11.009
38. Luxán G, Casanova JC, Martínez-Poveda B, Prados B, D'Amato G, MacGrogan D, et al. Mutations in the NOTCH pathway regulator MIB1 cause left ventricular noncompaction cardiomyopathy. Nat Med. 2013;19(2):193–201. doi:10.1038/nm.3046
39. Siguero-Álvarez M, Salguero-Jiménez A, Grego-Bessa J, de la Barrera J, MacGrogan D, Prados B, et al. A human hereditary cardiomyopathy shares a genetic substrate with bicuspid aortic valve. Circulation. 2023;147(1):47–65. doi:10.1161/CIRCULATIONAHA.121.058767
40. Arndt AK, Schafer S, Drenckhahn JD, Sabeh MK, Plovie ER, Caliebe A, et al. Fine mapping of the 1p36 deletion syndrome identifies mutation of PRDM16 as a cause of cardiomyopathy. Am J Hum Genet. 2013;93(1):67–77. doi:10.1016/j.ajhg.2013.05.015
41. Kramer RJ, Fatahian AN, Chan A, Mortenson J, Osher J, Sun B, et al. PRDM16 deletion is associated with sex-dependent cardiomyopathy and cardiac mortality: a translational, multi-institutional cohort study. Circ Genom Precis Med. 2023;16(4):390–400. doi:10.1161/CIRCGEN.122.003912
42. Billon C, Millat G, Goudal A, Malan V, Khraiche D, Wahbi K, et al. Dilated cardiomyopathy in patients with PRDM16 haploinsufficiency. J Mol Med (Berl). 2025;103(11-12):1417–1428. doi:10.1007/s00109-025-02586-1
43. Jongbloed JDH, Lekanne Deprez RH, Vatta M. Introduction to molecular genetics. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. Chapter 1, p. 3–26 (p. 8). doi:10.1007/978-3-030-45457-9_1
44. Sun B, Rouzbehani OMT, Kramer RJ, Ghosh R, Perelli RM, Atkins S, et al. Nonsense variant PRDM16-Q187X causes impaired myocardial development and TGF-β signaling resulting in noncompaction cardiomyopathy in humans and mice. Circ Heart Fail. 2023;16(12):e010351. doi:10.1161/CIRCHEARTFAILURE.122.010351
45. van Mil A, Neef K, Balk GM, Buikema JW, Sluijter JPG, Doevendans PAFM. Induced pluripotent stem cells. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. Chapter 26, p. 439–455 (p. 444). doi:10.1007/978-3-030-45457-9_26
46. Saudubray JM, Baumgartner MR, García-Cazorla Á, Walter J, editors. Inborn Metabolic Diseases: Diagnosis and Treatment. 7th ed. Berlin: Springer; 2022. Chapter 35, section 35.3.8 (Cardiolipin remodelling enzyme [TAZ] deficiency: Barth syndrome).
47. Pignatelli RH, McMahon CJ, Dreyer WJ, Denfield SW, Price J, Belmont JW, et al. Clinical characterization of left ventricular noncompaction in children: a relatively common form of cardiomyopathy. Circulation. 2003;108(21):2672–2678. doi:10.1161/01.CIR.0000100664.10777.B8
48. Brescia ST, Rossano JW, Pignatelli R, Jefferies JL, Price JF, Decker JA, et al. Mortality and sudden death in pediatric left ventricular noncompaction in a tertiary referral center. Circulation. 2013;127(22):2202–2208. doi:10.1161/CIRCULATIONAHA.113.002511
49. Askarinejad A, Lüscher TF, Lip GYH. Atrial fibrillation in patients with left ventricular non-compaction: incidence and associations with stroke, heart failure, and mortality. Cardiovasc Res. 2026;122(13):1761–1769. doi:10.1093/cvr/cvag156
50. Kawamura H, Ikawa M, Hirono K, Kimura J, Okuno T, Kawatani M, et al. Low-frequency maternal novel MYH7 mosaicism mutation in recurrent fetal-onset severe left ventricular noncompaction: a case report. Front Pediatr. 2023;11:1195222. doi:10.3389/fped.2023.1195222
51. Gati S, Chandra N, Bennett RL, Reed M, Kervio G, Panoulas VF, et al. Increased left ventricular trabeculation in highly trained athletes: do we need more stringent criteria for the diagnosis of left ventricular non-compaction in athletes? Heart. 2013;99(6):401–408. https://doi.org/10.1136/heartjnl-2012-303418 PMID: 23393084
52. Firth HV, Hurst JA. Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. Hypertrophic cardiomyopathy (HCM), p. 460–461.
53. Jones KL, Jones MC, del Campo M. Smith's Recognizable Patterns of Human Malformation. 8th ed. Philadelphia: Elsevier; 2022. 1p36 deletion syndrome, p. 76; Walker-Warburg syndrome, p. 256; Coffin-Lowry syndrome, p. 402; Beals syndrome, p. 666.
54. Scott DA, Hernandez-Garcia A, Azamian MS, Jordan VK, Kim BJ, Starkovich M, et al. Congenital heart defects and left ventricular non-compaction in males with loss-of-function variants in NONO. J Med Genet. 2017;54(1):47–53. doi:10.1136/jmedgenet-2016-104039
55. Singer JS, Garczarczyk-Asim D, Michel M, Müller T, Hackl L, Janecke AR. Heart transplant for noncompaction cardiomyopathy in NONO-related syndromic intellectual disability. Mol Genet Genomic Med. 2026;14(7):e70261. doi:10.1002/mgg3.70261
56. Bausch Health US LLC. VASOTEC® (enalapril maleate) tablets. Prescribing information. NDA 018998. Silver Spring (MD): U.S. Food and Drug Administration; SPL effective 2026-01-23. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=39631f1f-5d19-43c1-b504-bf56d991ed97 (accessed 2026-10-26)
57. Waylis Therapeutics LLC. COREG® (carvedilol) tablets. Prescribing information. NDA 020297. Silver Spring (MD): U.S. Food and Drug Administration; SPL effective 2026-07-31. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=07c619c6-4a0c-445e-94fa-664a54f68a39 (accessed 2026-10-26)
58. Melinta Therapeutics LLC. TOPROL-XL® (metoprolol succinate) extended-release tablets. Prescribing information. NDA 019962. Silver Spring (MD): U.S. Food and Drug Administration; SPL effective 2026-03-03. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=496ddfaa-7c9c-4888-b747-a210249b367a (accessed 2026-10-26)
59. Novartis Pharmaceuticals Corporation. ENTRESTO® (sacubitril and valsartan) tablets; ENTRESTO® SPRINKLE oral pellets. Prescribing information. NDA 207620/218591. Silver Spring (MD): U.S. Food and Drug Administration; SPL effective 2026-07-06. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=000dc81d-ab91-450c-8eae-8eb74e72296f (accessed 2026-10-26)
60. European Medicines Agency. Entresto (sacubitril/valsartan): EPAR – product information (summary of product characteristics). Amsterdam: EMA. Available from: https://www.ema.europa.eu/en/documents/product-information/entresto-epar-product-information_en.pdf (accessed 2026-10-26)
61. Pfizer Laboratories Div Pfizer Inc. ALDACTONE® (spironolactone) tablets. Prescribing information. NDA 012151. Silver Spring (MD): U.S. Food and Drug Administration; SPL effective 2025-11-28. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=0fed2822-3a03-4b64-9857-c682fcd462bc (accessed 2026-10-26)
62. AstraZeneca Pharmaceuticals LP. FARXIGA® (dapagliflozin) tablets. Prescribing information. NDA 202293. Silver Spring (MD): U.S. Food and Drug Administration; SPL effective 2026-06-03. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=72ad22ae-efe6-4cd6-a302-98aaee423d69 (accessed 2026-10-26)
63. European Medicines Agency. Forxiga (dapagliflozin): EPAR – product information (summary of product characteristics). Amsterdam: EMA. Available from: https://www.ema.europa.eu/en/documents/product-information/forxiga-epar-product-information_en.pdf (accessed 2026-10-26)
64. Validus Pharmaceuticals LLC. LASIX® (furosemide) tablets. Prescribing information. NDA 016273. Silver Spring (MD): U.S. Food and Drug Administration; SPL effective 2025-09-30. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=2c9b4d8f-0770-482d-a9e6-9c616a440b1a (accessed 2026-10-26)
65. U.S. Food and Drug Administration. 21 CFR 343.80 — Professional labeling (aspirin: comprehensive prescribing information, vascular indications). Electronic Code of Federal Regulations, version of 2026-10-23. Available from: https://www.ecfr.gov/current/title-21/chapter-I/subchapter-D/part-343/subpart-D/section-343.80 (accessed 2026-10-26)
66. Amgen Inc. NEUPOGEN® (filgrastim) injection. Prescribing information. BLA 103353. Silver Spring (MD): U.S. Food and Drug Administration; revised 2026-07 [SPL effective 2026-07-23]. Available from: https://dailymed.nlm.nih.gov/dailymed/lookup.cfm?setid=97cc73cc-b5b7-458a-a933-77b00523e193
67. European Medicines Agency. Zarzio (filgrastim): EPAR – product information (summary of product characteristics). Amsterdam: EMA. Available from: https://www.ema.europa.eu/en/documents/product-information/zarzio-epar-product-information_en.pdf
68. ClinicalTrials.gov. NCT07531251 — Phase 3b/4, randomized, double-blind, parallel-group, placebo-controlled trial to evaluate the efficacy and safety of daily subcutaneous injections of elamipretide in patients with genetically confirmed Barth syndrome (4TAZPower). Promotor: Stealth BioTherapeutics. Accessed on 7 October 2026. NCT07531251
69. Reid Thompson W, Hornby B, Manuel R, Bradley E, Laux J, Carr J, et al. A phase 2/3 randomized clinical trial followed by an open-label extension to evaluate the effectiveness of elamipretide in Barth syndrome, a genetic disorder of mitochondrial cardiolipin metabolism. Genet Med. 2021;23(3):471–478. doi:10.1038/s41436-020-01006-8
70. Thompson WR, Manuel R, Abbruscato A, Carr J, Campbell J, Hornby B, et al. Long-term efficacy and safety of elamipretide in patients with Barth syndrome: 168-week open-label extension results of TAZPOWER. Genet Med. 2024;26(7):101138. doi:10.1016/j.gim.2024.101138
71. ClinicalTrials.gov. Search for “noncompaction”, “non-compaction”, “hypertrabeculation” and “Barth Syndrome” (API v2) and registry entries NCT04424030, NCT06504056 and NCT04689360. Accessed on 7 October 2026. clinicaltrials.gov
72. Rouzbehani OM, Stephens S, Werbner B, Szulik MW, Sun B, Hua M, et al. Perm1 gene therapy mitigates PRDM16-associated cardiomyopathy [preprint]. bioRxiv. 2026. doi:10.64898/2026.04.13.718330
73. Nannenberg EA, Arens Y. Prenatal diagnosis and preimplantation genetic testing for inherited cardiac diseases. In: Baars HF, van der Smagt JJ, Doevendans PAFM, van Tintelen JP, editors. Clinical Cardiogenetics. 3rd ed. Cham: Springer; 2020. Chapter 27, p. 457–461 (p. 460). doi:10.1007/978-3-030-45457-9_27
74. Barth Syndrome Foundation. [www.barthsyndrome.org](https://www.barthsyndrome.org/)
75. Barth Syndrome Trust. [www.barthsyndrome.org.uk](https://www.barthsyndrome.org.uk/)
76. Children's Cardiomyopathy Foundation. [www.childrenscardiomyopathy.org](https://www.childrenscardiomyopathy.org/)
77. Cardiomyopathy UK. [www.cardiomyopathy.org](https://www.cardiomyopathy.org/)